Вакансии

Задачи: • Проектирование и разработка сложных Агентных и AgenticRAG систем. Обеспечение воспроизводимости и отказоустойчивости • Разработка решений начиная от прототипа, заканчивая выводом в промышленную эксплуатацию • Определение ML SysDes решения с учетом разрешенного технологического стека • Участие в валидации и автомониторинге моделей, проведение A/B тестирования Требования: • Высшее образование в сфере компьютерных наук, физ-мата, прикладной математики и информатики. Приветствуются дипломы ведущих технических и научных ВУЗов (МФТИ, МГУ, Сколтех, ВШЭ, ИТМО, МИФИ и др.) • Знание мат. статистики, классического ML, NLP, опыт в постановке и интерпретации экспериментов • Опыт e2e-разработки ML/AI-пайплайнов (в т.ч. RAG-пайплайнов) и проектировании архитектуры AI решений • Умение переводить бизнес-постановку задачи в ML-постановку • Высокий уровень владения Python и SQL • Знание фреймворков, библиотек, алгоритмов машинного обучения: XGBoost, CatBoost, PyTorch, TensorFlow, Transformers • Знание принципов построения RAG, VisionRAG и AgenticRAG систем • Знание архитектур нейронных сетей: RNN, LSTM, трансформеры (BERT, BART, T5) • Знание агентных архитектур (ReAct, Multi-agent и др.) • Знание фреймворков для работы с LLM (LangChain, LangGraph и др.) • Знание веб-фреймворков (FastAPI, Flask), контейнеризации (Docker) • Понимание жизненного цикла моделей (CRISP-DM) Условия: • Гибкий гибрид (обсуждаем индивидуально) • Современный IT-офис вблизи Москва-Сити с фитнес-залом • Ипотека с выгодой для сотрудника и льготные условия кредитования • Бесплатная подписка СберПрайм+ • Скидки на продукты компаний-партнеров • ДМС с первого дня и льготное страхование для близких • Корпоративная пенсионная программа • Обучение за счет компании: онлайн курсы в онлайн-школе Сбера и неограниченный доступ к библиотеке, обучение в Корпоративном университете, тренинги, митапы и возможность получить новую квалификацию • Крупнейшее DS&AI Community — более 600 DS банка, регулярный обмен знаниями, опытом и лучшими практиками, интерактивные лекции и мастер-классы от ведущих ВУЗов и экспертов технологических компаний, дайджест о самых последних разработках в области DS&AI и отчеты с крупнейших конференций мира

Location: Bremen, Germany Supervisor: Prof. Dr. Alexander Tormasov About the Position Constructor Knowledge Labs (CKL), in collaboration with Constructor Technology, invites applications for a PhD position in machine learning for software engineering and formal methods, on the Constructor Fabric project. Constructor Fabric turns a company's informal knowledge into production software through a pipeline of composable capability units called gears : requirements → architecture → product fit → framework (G1) → application → runtime. We are building a World Model over that pipeline — a model that does not just generate code, but predicts the consequences of an architectural decision: total cost of ownership, unintended side effects, latency and failure behaviour, and whether a proposed composition is even admissible. The central object is not a digital twin of the application but its formal architectural skeleton : gear contracts (GearSpec), a typed attributed hypergraph of the application (AppGraph), a composition algebra that defines which assemblies are legal, and several semantic projections (types, protocols, resources, security) over which properties are proved, refuted by counterexample, or reported as an explainable gap . All of it is written in a domain-specific language whose syntax trees and graphs we keep in a projectional editor (JetBrains MPS) — so the DSL is not a convenience layer, it is the thing that defines the model's state space and the boundary of what the model is allowed to propose. What you will work on Depending on your background, one or two of: Formal skeleton and composition — GearSpec as a contract $(Requires, Provides, State, Effects, Assumptions, Guarantees, Parameters)$; AppGraph as a typed attributed hypergraph; an algebra of assembly (sequential, parallel, feedback, refinement); semantic projections and system-level invariants over trajectories. Architecture synthesis as an inverse problem — given a specification, load and constraints, find the composition of gears that realises it. Constraint solving and MCTS as the baseline; a learned transition/value model for ranking and pruning; LLM proposals treated strictly as candidate hypotheses that a DSL parser, type checker and property checks must then admit or reject. DSL design and language engineering — the gear DSL and its glue layer decide which AppGraphs are expressible at all, and therefore where the search space ends. This thread covers the abstract syntax and static semantics of that DSL; the requirement that an architectural decision be unbypassable by construction (you must not be able to wire authentication around the Auth Resolver); what happens to the language when gear contracts change or a 201st gear is added; projectional representation of graphs with cycles; and the open hypothesis that a specialised G1-DSL is generated together with every G1. Local transition models and pipeline surrogates — fast approximate predictors for single pipeline steps (spec change → code quality, architecture → product fit, app → runtime outcome), with calibrated uncertainty, validated against real telemetry. Data — reconstructing (specification → architecture) pairs from open repositories at scale, plus synthetic project generation from a fixed gear set and DSL grammar to cover rare transitions, gaps and counterexamples. Profile MSc (or equivalent) in computer science, mathematics, or a mathematical natural science. Coursework in compilers, programming-language theory, language design or model-driven engineering is a strong positive signal, half of this project is deciding what a formal architecture description is allowed to say. Evidence that you finish things: a maintained open-source repository, a peer-reviewed publication, or a strong competition result. Fluent technical English. Most importantly: the habit of stating an assumption explicitly and designing an experiment that could falsify it. Required mathematical background Core skills: Probability and statistics. Estimation, Bayesian inference, uncertainty quantification and calibration (proper scoring rules, reliability diagrams, ECE), experiment design and significance testing. A prediction we cannot calibrate is not usable here. Linear algebra and multivariate calculus at the level needed to derive, not just call, a training objective: matrix calculus, gradients, stochastic optimisation, conditioning. Discrete mathematics and graph theory. Typed and attributed graphs and hypergraphs, subgraph matching and isomorphism, graph rewriting, reachability, basic complexity classes. Mathematical logic and formal methods. First-order logic, type systems, SAT/SMT, temporal logic (LTL/CTL) and model checking, pre/post-conditions, invariants, refinement. Sequential decision making. MDP/POMDP, dynamic programming, model-based RL, latent state-space (world) models, Monte-Carlo Tree Search. Strong command of at least one, working knowledge of a second: Formal languages, semantics and language engineering — grammars (context-free, attribute and graph grammars), derivation and reachability, parsing; abstract versus concrete syntax; operational and denotational semantics; metamodelling, model-driven engineering, model-to-model transformation, and projectional (structural) editing of the kind JetBrains MPS implements. We are asking for the education, not the tool: you should know why a projectional editor exists, what a metamodel buys you over a text grammar, and how a language evolves without breaking its existing models. Hands-on MPS experience is explicitly not required — we will teach it. Category theory for compositional systems — monoidal and traced monoidal categories, functors as semantic projections, string diagrams, operads. This is the language of our assembly algebra. Causal inference — structural causal models, do-calculus, counterfactuals, identifiability. Required for the "what happens if we replace this technology / move on-premises" class of questions, which similarity search alone cannot answer. Queueing theory and applied stochastic processes — M/M/1 and networks of queues, heavy tails, retries, timeouts, partial failure and reliability models. This is the resource and stochastic projection. Combinatorial optimisation, constraint programming and program synthesis theory — CEGIS, version-space algebras, symmetry breaking, search heuristics. Representation learning theory — VAEs and latent dynamics, GNN expressiveness (Weisfeiler-Leman hierarchy), permutation equivariance, embedding migration under a changing vocabulary. Required hands-on programming skills This is an engineering-heavy PhD. We expect code that another person can clone, run and reproduce — not notebooks that worked once. Must have: Python (professional level). Type hints, packaging, `pytest`, profiling, `asyncio`, memory-aware data structures. You write unbuffered, logged, restartable long-running jobs by default. Training loops written from scratch, custom datasets and collation, mixed precision, multi-GPU; experiment tracking (W&B or MLflow), configuration management (Hydra), strict seed and artefact reproducibility. Graph machine learning. PyTorch Geometric or DGL — message passing, heterogeneous and typed graphs, graph transformers, handling graphs whose node vocabulary changes between versions. Solvers and verification tooling driven from Python. Z3 or cvc5 hands-on; demonstrated exposure to at least one model checker (TLA+/TLC, SPIN, NuSMV, Alloy) or proof assistant (Lean, Coq, Isabelle). Code as data. AST manipulation, `tree-sitter`, static analysis, dependency and interface graph extraction, large-scale repository mining. Parallel data pipelines. `multiprocessing`, Ray or Dask; checkpointing, heartbeats, cost estimation before launching a run that touches millions of files. LLM engineering. API orchestration, structured and grammar-constrained decoding, evaluation harnesses, token/cost accounting — with the discipline to treat model output as a hypothesis subject to formal checks. Linux, git, code review, shell. Desirable: Language tooling you have actually used: ANTLR, Xtext, Lark, or a parser, type checker or interpreter you wrote yourself. Our internal representation of gear syntax trees and graphs lives in JetBrains MPS — again, we do not ask for MPS experience, only that you are comfortable reading and extending a metamodel; Java/Kotlin literacy helps once you get there. JAX, or compiler / SAT-SMT internals. Infrastructure literacy — Docker, Kubernetes, Terraform, OpenTelemetry/Prometheus — since the resource projection is calibrated against real deployment telemetry. One systems language (Rust, Go or C++) for performance-critical search. Prior experience with production software architecture, or with a brownfield system you had to change without breaking. Funding The appointment provides full financial coverage through a dedicated fellowship, comprising: Monthly stipend of €1,650 Monthly research-cost allowance of €100 (Forschungskostenpauschale) Health-insurance subsidy of €100 per month Location: Bremen, Germany Duration: 3 years, full time Start date: November 2026 Application package must include CV, including publications, code, or research projects. Cover letter (1–2 pages) explaining your motivation, research interests, and fit with the project. A link to code you have written. A one-page research statement picking one open research question from the project and sketching how you would approach it. Master’s and Bachelor’s degree certificate and transcripts. Contact information for 2–3 academic referees. Applications are reviewed on a rolling basis. Shortlisted candidates will be invited to interviews.

Location: Remote - United States Reddit is a community of communities. It’s built on shared interests, passion, and trust, and is home to the most open and authentic conversations on the internet. Every day, Reddit users submit, vote, and comment on the topics they care most about. With 100,000+ active communities and approximately 130 million daily active unique visitors, Reddit is one of the internet’s largest sources of information. For more information, visit www.redditinc.com . The AI Security team within Reddit’s Security Platform Engineering organization builds security into Reddit’s products, engineering systems, AI platforms, and operational infrastructure so the secure path is the easiest path for both people and agents. A core part of this work is developing practical, high-quality machine learning systems that detect and prevent risks such as prompt injection, jailbreaks, sensitive data exposure, and unsafe or unauthorized AI behavior. Building on Reddit’s centralized LLM Guardrails Platform, which provides a shared security and safety boundary across major products, we are expanding ML-powered protections as Reddit’s AI systems and threat landscape evolve. We are looking for a founding Staff Machine Learning Engineer to lead the development, training, and optimization of models for AI security at Reddit. This is a strategic and hands-on individual contributor role, combining deep ownership of model architecture, training data, and experimentation with technical leadership across teams. You will help Reddit deliver stronger AI protections while preserving a high-quality product experience. How You’ll Have Impact Select, adapt, fine-tune, evaluate, and deploy pretrained models and lightweight classifiers for Reddit-specific security problems. Build reproducible training and evaluation pipelines on Reddit’s ML platform, partnering with platform engineers to improve inference performance, resource efficiency, and operational reliability. Set the technical vision and multi-quarter modeling roadmap, partnering with cross-functional teams to gather requirements, define model architectures, and iterate on model development. Conduct model evaluations and performance analysis to improve accuracy and adversarial robustness, and define launch criteria that balance false positives, latency, throughput, reliability, and cost. Own training-data quality and the production model lifecycle, using monitoring, incident findings, and red-team feedback to guide dataset improvements, retraining, and safe rollout or rollback. Establish best practices for responsible ML development and deployment, including reproducible experiments, testing, model and data lineage, and privacy-aware data use. Stay current with research in NLP, large language models, and relevant multimodal techniques, translating promising advances into measurable model improvements. Mentor engineers and lead technical discussions and reviews, shaping the team’s long-term ML capabilities and AI security direction. Who You Might Be 8+ years of experience developing machine learning models, with substantial hands-on model training experience, demonstrated production impact, and a record of leading complex initiatives across teams. Strong background in Python programming, software engineering, and deep learning frameworks and libraries such as TensorFlow, PyTorch, or Hugging Face Transformers. Deep understanding of neural network architectures and optimization, with proficiency in data preprocessing, tokenization, embeddings, language modeling, and model calibration. Expertise in scalable data pipelines and distributed training frameworks such as Ray Train or PyTorch Distributed, with a strong understanding of hardware and system tradeoffs. Demonstrated rigor in experimental design and model evaluation, including representative holdouts, ablation studies, adversarial tests, and error analysis to diagnose training issues, bias, and generalization gaps. Excellent written and verbal communication, with the ability to explain model behavior, security risk, uncertainty, and tradeoffs to technical and non-technical partners. Preferred Qualifications Experience applying ML to security, trust and safety, fraud, privacy, or related adversarial domains. Experience with adversarial training, model distillation, active learning, or synthetic-data generation to improve model quality, robustness, and training efficiency. Benefits: Comprehensive Healthcare Benefits and Income Replacement Programs 401k with Employer Match Global Benefit programs that fit your lifestyle, from workspace to professional development to caregiving support Family Planning Support Gender-Affirming Care Mental Health & Coaching Benefits Flexible Vacation & Paid Volunteer Time Off Generous Paid Parental Leave #LI-Remote Pay Transparency: This job posting may span more than one career level. In addition to base salary, this job is eligible to receive equity in the form of restricted stock units, and depending on the position offered, it may also be eligible to receive a commission. Additionally, Reddit offers a wide range of benefits to U.S.-based employees, including medical, dental, and vision insurance, 401(k) program with employer match, generous time off for vacation, and parental leave. To learn more, please visit https://www.redditinc.com/careers/ . To provide greater transparency to candidates, we share base salary ranges for all US-based job postings regardless of state. We set standard base pay ranges for all roles based on function, level, and country location, benchmarked against similar stage growth companies. Final offer amounts are determined by multiple factors including, skills, depth of work experience and relevant licenses/credentials, and may vary from the amounts listed below. The base salary range for this position is: $230,000 — $322,000 USD In select roles and locations, the interviews will be recorded, transcribed and summarized by artificial intelligence (AI). You will have the opportunity to opt out of recording, transcription and summarization prior to any scheduled interviews. During the interview, we will collect the following categories of personal information: Identifiers, Professional and Employment-Related Information, Sensory Information (audio/video recording), and any other categories of personal information you choose to share with us. We will use this information to evaluate your application for employment or an independent contractor role, as applicable.  We will not sell your personal information or disclose it to any third party for their marketing purposes.  We will delete any recording of your interview promptly after making a hiring decision.  For more information about how we will handle your personal information, including our retention of it, please refer to our Candidate Privacy Policy for Potential Employees and Contractors . Reddit is proud to be an equal opportunity employer, and is committed to building a workforce representative of the diverse communities we serve.  Reddit is committed to providing reasonable accommodations for qualified individuals with disabilities and disabled veterans in our job application procedures. If, due to a disability, you need an accommodation during the interview process, please let your recruiter know.

Location: New York, NY (HQ) Workplace: Hybrid ABOUT RAMP Ramp is building the smart infrastructure for finance teams, embedded in the transaction flow of every dollar a business spends. We automate how over $275B in annualized spend flows in and out of 85,000+ companies: authorizing payments, flagging risk, categorizing spend, and closing books. The problems are high-stakes, data-dense, and unforgiving. We hire people with high agency and high urgency. We look for slope over intercept. We care less about where you trained and more about what you’ve built. At Ramp, everyone is a builder who owns problems end to end and makes consequential decisions that shape the outcome. The median Ramp customer saves 7.5% and grows revenue 16% in their first year – far in excess of businesses operating without Ramp. We believe every ambitious company deserves the same. If you want to build systems that directly shape how companies move and manage billions, Ramp is the place to do it. ABOUT THE ROLE The Applied Science team builds models and tools that solve Ramp’s most critical problems: from underwriting businesses to combatting fraud to making spend management smarter. We’re deeply embedded in the business and provide a quantitative foundation for decision making. As an Applied Science intern, you’ll be a fully integrated member of the team and own your project from start to finish. Working with engineers, product managers, and business stakeholders, you’ll translate complex business needs into scalable machine-learning-driven solutions. This is a chance to apply ML concretely, ship code, and create genuine value for Ramp and our customers. You will focus on exciting problems in areas like: credit, fraud, growth, or our core product.   WHAT YOU’LL DO - End-to-End ML: own the model lifecycle from data exploration and feature engineering to training, benchmarking, deployment, and monitoring - State-of-the-Art AI: leverage the latest Large Language Models (LLMs) to solve novel problems and create new product capabilities for our customers - Versatile Techniques: apply the right tools to the right problems, whether it’s deep learning, gradient boosting, or causal inference - Rigorous Experimentation: quantify the impact of your work through A/B tests and other statistical methods - Collaborate: partner closely with product and business leaders to translate models and insights into actionable strategy and user-facing features   WHAT YOU NEED - B.S., M.S. or Ph.D. Student: currently pursuing a degree in Data Science, Computer Science, Math, Physics, Economics, Statistics, or other quantitative fields with an expected graduation date between Dec 2027 - 2029. Graduate degrees are preferred, but not a must. - Strong ML Fundamentals: solid understanding of the mathematical foundations of machine learning, statistics, probability, and optimization - Strong Interest or Experience with AI: curiosity and drive to integrate cutting edge LLMs and agents into applied solutions - Python Proficiency: good grasp of common Data Science libraries (pandas, scikit-learn, NumPy, PyTorch, etc.) - SQL Knowledge: experience wrangling data in a modern data warehouse (e.g. Snowflake, BigQuery, Redshift, Clickhouse) - Practical Experience: track record of curating datasets and building/evaluating ML models - Strong Communication: ability to clearly explain complex concepts to both technical and non-technical audiences and use data to build a compelling narrative - Bias For Action: a comfort with ambiguity and desire to ship solutions quickly then iterate   NICE TO HAVES - Publications, Projects, or Previous Experience: relevant experience applying AI/ML and demonstrating your passion for the field - Production ML Mindset: knowledge of software engineering best practices applied to ML including version control (Git), testing, and writing maintainable code - Data Orchestration: experience with leveraging modern data orchestration platforms (Airflow, Dagster, Prefect, Metaflow)   COMPENSATION - The monthly rate for this internship is $12,500 USD + housing stipend   RAMP BENEFITS - Apple MacBook - Catered lunches in NYC office Monday-Friday - Weekly coffee stipend BENEFITS AVAILABLE TO ALL FULL-TIME RAMP EMPLOYEES (GLOBAL) - Flexible PTO - Centralized home-office equipment ordering - Health and wellness stipend - Budget for intra-office travel - Weekly coffee stipend UNITED STATES - 100% medical, dental & vision insurance coverage for you, with partial coverage for dependents - One Medical annual membership - 401(k), including employer match on contributions made while employed by Ramp - Fertility HRA (up to $10,000 per year) - Parental leave: up to 16 weeks (birthing + bonding) or 8 weeks (bonding only) at 100% pay - Pet insurance - In-office perks: lunch, snacks, drinks, and more - Relocation expense coverage to NYC or SF (if needed) CANADA - Group medical, dental, and vision coverage through Sun Life - Life, AD&D, and disability coverage - Fertility drug coverage (up to $4,000 lifetime) - Group Retirement Plan with employer match (RRSP + DPSP) - Parental leave: up to 16 weeks (birthing + bonding) or 8 weeks (bonding only) at 100% pay, with additional time available at reduced pay - Employee Assistance Program and virtual care through Lumino Health UNITED KINGDOM - Private medical insurance through Freedom Elite - Virtual GP and at-home care via eMed x Livi - Workplace pension through Penfold, with salary sacrifice option - Parental leave: up to 16 weeks (birthing + bonding) or 8 weeks (bonding only) at 100% pay with additional time available at reduced pay REFERRAL INSTRUCTIONS If you are being referred for the role, please contact that person to apply on your behalf.   OTHER NOTICES Pursuant to the San Francisco Fair Chance Ordinance, we will consider for employment qualified applicants with arrest and conviction records.   Beware of recruiting scams: Ramp will only contact you through official @Ramp.com http://Ramp.com email addresses and will never ask for payment or sensitive personal information during the hiring process.   Ramp Applicant Privacy Notice https://ramp.com/legal/applicant-privacy-notice

Задачи: — разработка и внедрение LLM моделей для решения NLU задач бизнеса (SFT, RAG (Retrieval Augmented Generation), Agents, Summarization, Text Ranking, Text Matching, Language Modeling) — дообучение LLM моделей — разработка подходов и процессов разметки данных для оценки качества работы LLM — формирование и работа с ML пайплайнами: работа с данными, обучение/дообучение NLP моделей, оценка качества решений, поддержка/автоматизация решений — оптимизация работы моделей для промышленного контура на CPU/GPU — работа с командой бизнес-представителей, DS-разработчиками — организация проверки и генерация гипотез для решения технических и бизнес-задач. Требования: — опыт разработки на python, numpy, sklearn, pandas + библиотеки обработки текстовых данных — опыт работы с Pytorch для построения DL текстовых моделей — опыт работы с библиотеками LangChain/LangGraph — опыт практической работы с LLM через API — отличные теоретические знания классического и нейросетевого NLP, в тч LLM — опыт дообучения классических трансформеров и LLM — практический опыт, эксперименты, внедрение в ПРОД LLM решений — опыт prompt-engineering Будет плюсом: — работа с инструментами Hadoop (HDFS, Hive), Spark — опыт постановки и проведения a/b тестов — опыт работы с распределенным обучением, глубокое знание GPU архитектуры. Условия: — комфортный современный офис рядом с м. Кутузовская — возможен гибридный формат работы — ежегодный пересмотр зарплаты, квартальная и годовая премия — корпоративный спортзал и зоны отдыха более 400 образовательных программ СберУниверситета для профессионального и карьерного развития — программа адаптации и помощь руководителя на старте — расширенный ДМС, льготное страхование для семьи и корпоративная пенсионная программа — гибкий дисконт по ипотечному кредиту, равный 1/3 ключевой ставки ЦБ — подписка Прайм с возможностью совместного использования на трёх близких — вознаграждение за рекомендацию друзей в команду Сбера.

Location: Paris ABOUT MISTRAL Mistral provides full-stack AI solutions: from frontier models to developer tools, applications, and compute. We partner with enterprises tackling the hardest problems across high-stakes industries like finance, manufacturing, defense, healthcare, and the public sector, co-creating customized AI systems that they can run on their terms. We are a dynamic, collaborative team passionate about AI and its potential to transform society. Our diverse workforce thrives in competitive environments and is committed to driving innovation. Our teams are distributed between Europe, North America, Asia and the Middle East. We are creative, low-ego and team-spirited. The Role   As an Applied Engineering Intern, you will work closely with our Applied AI Engineering team to facilitate the adoption of Mistral AI products among customers and collaborate with them to address complex technical challenges. This role is based in Paris, with an internship duration of 6 months. We are open to CIFRE programs as a continuation after the internship.   What you will do   • Contribute to the deployment of state-of-the-art GenAI applications, driving technological transformation with our customers. • Collaborate with researchers, AI engineers, and product engineers on complex customer projects. • Work with the product and science team to continuously improve our product and model capabilities based on customer feedback.   How We Work in Applied AI   - We care about people and outputs. - What matters is what you ship, not the time you spend on it - Bureaucracy is where urgency goes to vanish. You talk to whoever you need to talk to. The best idea wins, whether it comes from a principal engineer or someone in their first week. - Always ask why. The best solutions come from deep understanding, not from copying what worked before - We say what we mean. Feedback is direct, timely, and given because we care. - No politics. Low ego, high standards. - We embrace an unstructured environment and find joy in it.   About you   • You are currently pursuing a degree in AI, data science, or a related field from a tier 1 engineering school or university. • You have strong programming skills in Python. • You are familiar with machine learning algorithms and natural language processing techniques. • You hold basic understanding of MLOps and deploying machine learning use cases. • You have good communication skills with the ability to explain technical concepts to both technical and non-technical audiences.   Ideally you have: • Experience with deep learning frameworks such as PyTorch. • Familiarity with version control systems (e.g., Git) and Linux shell environment. • Experience working in HPC Environments • Publication record in AI or a related field Location & work authorization: This internship is available only in France. Please apply only if you have the required work authorization to complete the internship in the relevant country for its full duration.   WHAT WE OFFER We offer a comprehensive benefits package designed to support your well-being, growth, and work-life balance. Benefits vary by country and may include healthcare coverage, parental leave, retirement plans, relocation support, wellness programs, meal and transportation allowances, and other location-specific perks. For the most up-to-date details on benefits available in your location, please refer to our Benefits page https://app.notion.com/p/mistralai/Benefits-at-Mistral-36e6ba59a7fe836b93dd01737fcc27ef?source=copy_link. PRIVACY POLICY Your privacy matters to us. You can learn more about how we handle your personal data in our Applicant Privacy Policy https://legal.mistral.ai/terms/applicant-privacy-policy.

Location: Москва Наша команда отвечает за **независимую оценку и контроль качества генеративных моделей Банка.** Наша основная задача оценить флагманские генеративные модели Сбера – GigaChat, Kandinsky, GigaCode и PhysicalAI решений. Мы первыми видим, на что способна новая модель и проводим оценку новых возможностей прежде чем она выйдет к пользователям. Мы не просто «проверяем метрики», а: * глубоко разбираемся в подходах команд-разработчиков и в бизнес-процессах, где работают модели * критически рассматриваем архитектурные и методологические решения и предлагаем улучшения * разрабатываем собственные проверки, метрики и бенчмарки и фреймворки для их запуска * проводим оценку логов пользователей для построения наиболее релевантного тестирования Отдельный фокус – R&D деятельность: * создание новых методологий оценки качества * участие в разработке и развитии передовых банковских и российских бенчмарков (MERA, POLLUX и др.) * адаптация передовых мировых бенчмарков под специфику банковских кейсов * и многое другое Эта роль про независимое измерение фундаментального качества модели: не про их обучение – этим занимается отдельная команда Обязанности **Придумывать, как вообще измерить модель** ИИ стремительно развивается в мире и выходят новые модели и их виды, однако не всегда любая новая модель стала лучше предыдущей. Вам будет необходимо создать план по её проверке: что именно измерять, на какой выборке, с чем сравнивать, какая разница вообще будет считаться разницей в реальном качестве. **Отвечать за достоверность измерений** Полученные метрики не всегда отражают только лишь качество модели. Нужно уметь отделять реальный прирост качества от прочих факторов: ловить утечки данных в обучение, отличать дефект модели от дефекта замера, промпта или обвязки и проявлять тот же скепсис к собственной методологии: измеряем ли мы то, что заявили. **Разбираться в причинах ошибок** «Модель ошиблась» – это не вывод, а начало работы. Найденная вами причина той или иной проблемы модели превращается в рекомендацию команде разработки, а рекомендация – в следующую версию модели, которая этой ошибки уже не делает. **Строить автоматическую оценку** Разрабатывать LLM- и VLM-судей под конкретные задачи, разрабатывать с нуля собственные или дорабатывать существующие библиотеки валидации полного цикла и следим за фронтиром AI-исследований в области оценки моделей. **Строить то, чем пользуются другие** Фреймворки оценки, который вы развиваете – общие для команд по всему Банку. Бенчмарки, которые вы делаете, являются достоверной базой реального качества для разработчиков Банка и вне. Требования * **Умение проводить технический и научный ресерч**: находить и критически разбирать актуальные статьи, бенчмарки, модели и реализации; переводить ресерч в конкретные идеи для тестирования и развития внутренних подходов. * **Статистика и дизайн эксперимента:** доверительные интервалы, размер выборки, проверка гипотез, поправки на множественные сравнения * **Понимание современных AI-агентов и агентных систем**: понимать, чем агент отличается от обычного вызова LLM; знать основные компоненты агентной системы: роли, инструменты, контекст, память и оркестрация; * **Понимаете, как устроены генеративным модели изнутри:** трансформер и механизм внимания, диффузии, токенизация, параметры генерации, этапы обучения * **Опыт с генеративными моделями и понимание, как их оценивают:** метрики, бенчмарки, human evaluation, Model-as-a-judge и его ограничения * **Продуктовое мышление**: связывать тестирование с реальными пользовательскими сценариями и понимать, какие ошибки критичны для продукта, понимать влияние результатов на релиз и дальнейшее развитие модели. * **Python на продакшн-уровне:** чистый код по PEP 8, ООП и типовые паттерны проектирования, аннотации типов, привычка к линтерам и статическому анализу (ruff, mypy). Создаете и поддерживаете production-код eval-пайплайнов, читаете и ревьюите чужой. **Будет плюсом:** R&D-опыт или участие в создании бенчмарков; мультимодальные модели (изображения, видео, аудио); evaluation-фреймворки и harness; RAG; работа с асессорами; coding-агенты (Claude Code, Codex, Cursor); опыт работы с оркестрацией: LangGraph, OpenClaw, навыки вайбкодинга и AI-assisted coding **Стек:** Python, pandas, NumPy, PyTorch, HuggingFace (transformers, datasets), vLLM, SGLang, инференс-API, S3, Git, Docker, CI/CD, Linux, Bash. Условия * комфортный современный офис рядом с м. Кутузовская * гибкий формат и удаленной работы, в т.ч. на территории РФ * ежегодный пересмотр зарплаты и годовая премия * корпоративный спортзал и зоны отдыха * более 400 образовательных программ СберУниверситета для профессионального и карьерного развития * расширенный ДМС, льготное страхование для семьи и корпоративная пенсионная программа * гибкий дисконт по ипотечному кредиту, равный 1/3 ключевой ставки ЦБ * подписка Прайм с возможностью совместного использования на трёх близких * скидки на продукты компаний-партнеров * вознаграждение за рекомендацию друзей в команду Сбера.

Задачи: — Проектировать и разрабатывать multi-agent системы. — Интегрировать LLM с внутренними инструментами и сервисами. — Разрабатывать backend-сервисы и API для agent workflows. — Реализовывать RAG pipelines и tool/function calling. — Создавать CLI-based agent workflows и инструменты автоматизации. — Разрабатывать Computer Vision pipelines: object detection, multi-object tracking, person re-identification. — Разрабатывать multi-camera multi-person tracking для сетей камер. — Работать с видеопотоками в реальном времени и оптимизировать GPU inference. Требования: — Опыт работы с Node.js ecosystem и Express.js. — Уверенное владение Python. — Опыт работы с LLM APIs и prompt engineering. — Опыт работы с RAG, tool calling или agent frameworks. — Знакомство с Docker, Linux и Git. — Опыт работы с PyTorch и OpenCV. — Опыт в object detection и multi-object tracking. — Опыт работы с YOLO, ByteTrack / BoT-SORT или аналогами. — Понимание person re-identification и vector search. — Опыт работы с Milvus / FAISS. — Опыт обработки видео с ffmpeg / GStreamer. — Работа с RTSP и hardware decoding. Будет плюсом: — Опыт в NLP и ML. — Опыт развёртывания open-source LLM. — Опыт оптимизации inference. — Опыт работы с Vision-Language Models. — TensorRT / ONNX Runtime. Условия: — Удалённый формат работы. — Заработная плата: 1.600–3.200 USD net в месяц.

Задачи: — разработка моделей с применением алгоритмов и техник машинного обучения; — развёртывание моделей в выбранной среде; — обработка и устранение возникающих проблем, связанных с работой модели после вывода в продуктивное использование. Требования: — опыт работы по направлению не менее 2 лет; — знание алгоритмов и техник машинного обучения, их преимуществ, недостатков и особенностей применения; — понимание теории вероятности и математической статистики; — опыт программирования на языке Python, уверенное владение синтаксисом и основными конструкциями языка, использование библиотек для анализа данных и машинного обучения, знание библиотек для работы с веб-технологиями (Numpy, Pandas, matplotlib, seaborn, scikit-learn, XGBoost, CatBoost, PyTorch); — знание системы контроля версий и методов CI/CD; — понимание принципов работы с API и опыт их интеграции в проекты. Условия: — удалённый формат работы либо гибрид. — ДМС со стоматологией, индексация зарплаты, бонусы за командные и личные достижения, кафетерий льгот (20 000 на услуги от образования до здоровья каждый год), программы финансовой, психологической и юридической поддержки, компенсация питания, программы адаптации. — индивидуальные планы развития, помощь в наращивании экспертизы, работа с внутренним кадровым резервом внутри Северстали и Севергрупп. — корпоративная культура, взаимоуважение и поддержка на всех уровнях управления.

Задачи: — Перевод исследовательских DS рабочих процессов (например, прокси-метрики, модели устаревания) в производственные ML пайплайны с использованием Airflow, WandB и Ray, создание повторно используемых шаблонов для других команд. — Применение и внедрение методов причинно-следственного вывода с использованием производственного ML стека (propensity scoring, IPW, TMLE) для решения важных задач измерения за пределами экспериментальных возможностей. Создание инструментов самообслуживания для неэкспертов для получения строгих причинных выводов в масштабе. — Сотрудничество с ML инженерами и продуктовыми командами для выявления возможностей улучшения инструментов, метрик и методов измерения, что приводит к значительному улучшению качества моделей и бизнес-результатов. — Использование богатых метаданных и сигналов вовлеченности Pinterest для построения дата-дривен фреймворков, от важности признаков до деиндексации контента, для повышения эффективности и скорости платформы. — Проектирование и создание централизованных инструментов ML платформы для улучшения создания, оценки и доверия к признакам и моделям, включая производственные системы, работающие ежедневно в масштабе для всех моделей. Требования: — Опыт работы не менее 2 лет в роли прикладного ученого данных, ML инженера, исследователя или разработчика ПО с значительным опытом внедрения ML в производство. — Сильные навыки Python; опыт работы с PyTorch или аналогичными фреймворками глубокого обучения; знакомство с распределенными вычислениями (Spark, Ray), предпочтительно Ray. — Энтузиазм к созданию инструментов и платформ, которые увеличивают влияние всей ML организации, а не решают разовые задачи. — Глубокие знания теории ML с сильными фундаментальными основами для анализа моделей ML с первых принципов. — Владение лучшими практиками разработки ПО, включая контроль версий, код-ревью и воспроизводимые ML пайплайны. — Опыт работы с инструментами управления рабочими процессами (Airflow, Prefect, Jenkins или аналогичными) для надежной оркестрации ML пайплайнов. — Степень бакалавра или магистра в релевантной области, например, компьютерных наук, или эквивалентный опыт. Условия: — Работа удаленная с необходимостью присутствия в офисе 3-5 раз в квартал для личного взаимодействия. — Позиция не предусматривает помощь с релокацией. Дополнительно: — Компания предоставляет равные возможности для всех кандидатов без дискриминации. — Зарплатный диапазон от $114,297 до $235,319 в год, также возможна доля в компании. — Работа в США, предпочтительно в Пало-Альто, Калифорния, или удаленно по США.

Задачи: — Владеть проблемой, превращая цели в AI-решения в продакшене. — Строить и адаптировать большие языковые модели (LLM); дообучать или дистиллировать модели для повышения производительности. — Обучать собственные модели, включая сбор и оценку данных. — Принимать решения о разработке или покупке на основе чётких метрик и ограничений. — Выпускать функции в продакшен и мониторить их эффективность. Требования: — Опыт инженерной работы 5+ лет с сильными навыками программирования. — Практический опыт с LLM: retrieval, использование инструментов и контекстное проектирование. — Подтверждённый опыт обучения моделей на реальных производственных данных. — Владение Python, PyTorch и использованием GPU. — Английский язык на уровне Intermediate и выше. Желательно: — Опыт работы с инференсом в реальном времени, модерацией контента или рекомендательными системами. — Знакомство с Docker, Kubernetes и облачными платформами. Условия: — Конкурентная зарплата и опционные программы — все являются партнёрами. — Официальное трудоустройство на Кипре (поддержка визы для сотрудника и супруга). — Комфортный офис в Лимассоле с обеспечением питания и перекусов. — Медицинская страховка для сотрудника и семьи. — Возможности для профессионального роста, уроки греческого языка и командные мероприятия.

Location: New York City, New York What you’ll be part of Garner is on a mission to transform the U.S. healthcare system — and we’re the only proven player doing exactly that. We partner with employers to redesign how healthcare works: applying 550+ proprietary clinical metrics across 80+ specialties to a dataset of 320M+ patients to identify the best-performing doctors, then using compelling incentives to steer members to the care that helps them get healthier, faster. The result is a rare “win win” — better care and lower costs for both members and employers. In just five years, our work has helped over 2.5 million people access higher-quality care and saved $1B in healthcare costs. We recently raised our Series E and have doubled five years running. If you've ever wanted your work to solve a problem that touches every person in this country, this is the opportunity to do exactly that. You'd be joining a team fundamentally reimagining healthcare in the U.S. — and using AI to scale that impact further and faster than anyone else can. Who we want This is a role for someone who wants to work on a hard problem, deliver real impact and be pushed to be the best version of themselves. As an Associate Applied Scientist, you’ll be embedded in the Applied Science team. Our members rely on us to answer hard questions — Which doctor should I see? What will it cost? When should we reach out, and how? — and the quality of those answers is determined by the algorithms behind them. This is not a dashboards or descriptive-analytics role. You will own production systems end-to-end: framing the problem, defining the objective function, choosing the right approach (ML, optimization, heuristics, expert systems, or a hybrid), shipping it, and improving it against real-world outcomes. The closest analog outside healthcare is a quantitative researcher at a top hedge fund. What you will do Ship production algorithmic systems end-to-end, from problem framing to launch to iteration Frame messy, real-world healthcare and business constraints into clear objectives, tradeoffs, and decision frameworks Define the set of metrics needed to judge whether a solution is working, and validate solutions before they ship Choose the right approach for each problem, from machine learning to optimization to heuristics to simple rules, based on what the problem actually calls for Bring new ideas to your work, experimenting with approaches beyond the obvious to get better results Produce rigorous, well-validated work others can rely on, and review your peers' code and analyses Build a deep understanding of the healthcare economy and Garner's place in it To make the role concrete, here are two problems on our near-term roadmap: Provider tiering optimization. Build a tiering algorithm that jointly optimizes geographic access and total-cost-of-care savings across our doctor network. The objective function, constraints, and tradeoff surface are all open design questions. Member engagement model. Build an ML system that ingests claims data and in-app behavior to choose the right channel and moment for each touchpoint — SMS, push, phone, or email — to influence member behavior toward better-quality, lower-cost care. What we're looking for A Bachelor's or Master's degree (in progress) in Computer Science, Mathematics, Statistics, Economics, or a related STEM field, with an expected graduation date of Fall 2026 or Spring 2027. A bias toward action, quickly translating ideas into working prototypes to test approaches. Strong applied problem-solving skills, with the ability to define good metrics and then deliver solutions that improve them. A solid command of the methods your work calls for and the data your models depend on. Strong judgment in choosing between statistical models, heuristics, optimization approaches, and simpler algorithmic methods depending on the problem. Strong communication skills, with the ability to present your work clearly to senior stakeholders. A desire to do work that matters: and to be held to a high standard while doing it. You're energized by a mission with measurable outcomes, not just good intentions. You engage with AI tools as part of how you do your best work, and hold yourself accountable to outcomes that matter beyond your team’s metrics. What you’ll get here You’ll work on problems that matter, at a company working to change healthcare at scale. You’ll work at the intersection of AI and systemic healthcare reform, where the problems we solve are as interesting and compelling as the mission. At Garner, you’ll take on real, ambitious problems with real ownership and autonomy, alongside exceptional, principles-based people who genuinely want you to win. It’s demanding by design. You’ll be challenged to stretch beyond what you thought possible and receive consistent coaching to help you grow and do the best work of your career. This isn't the right fit for everyone, and that's intentional. The people here are driven by what's at stake for real people, and that's what gives our intensity its purpose. Technologies we use Python, SQL, AWS, Snowflake, pandas, XGBoost, PyTorch, HuggingFace, modern LLM tooling and eval frameworks. We pick tools based on the problem, not the resume — bring your judgment. Please note: we are unable to sponsor or take over sponsorship of an employment visa at this time. Compensation The target base comp range for this role is $150,000. Individual compensation will depend on qualifications, skills, and applicable laws. In addition to base compensation, this role is eligible to participate in our equity incentive plan and benefits package, including: flexible PTO, Medical/Dental/Vision plan options, 401(k) with company match, flexible spending accounts, Teladoc Health, and more. Few roles offer this combination: work that is genuinely hard, a mission that is genuinely important, and a team that is genuinely invested in your success. If that’s what you’re looking for, we’d love to meet you. Fraud and Security Notice: Please be aware of recent job scam attempts. Our recruiters use getgarner.com and garnerhealth.com email domains exclusively. If you have been contacted by someone claiming to be a Garner recruiter or a hiring manager from a different domain about a potential job, please report it to law enforcement here and to candidateprotection@garnerhealth.com . Equal Employment Opportunity: Garner Health is proud to be an Equal Employment Opportunity employer and values diversity in the workplace. We do not discriminate based upon race, religion, color, national origin, sex (including pregnancy, childbirth, reproductive health decisions, or related medical conditions), sexual orientation, gender identity, gender expression, age, status as a protected veteran, status as an individual with a disability, genetic information, political views or activity, or other applicable legally protected characteristics. Garner Health is committed to providing accommodations for qualified individuals with disabilities in our recruiting process. If you need assistance or an accommodation due to a disability, you may contact us at talent@garnerhealth.com .

Location: Hong Kong Nex is on a mission to help families rediscover the joy of movement. Created by parents for parents, Nex combines technology and play to deliver fun, social, and interactive experiences powered by natural body motion, encouraging kids and adults to move more, play more, and have fun together. Nex Playground, the company’s award-winning active play system, is purpose-built to get families moving year-round, with safety and privacy as core considerations in its intentional design. It is certified kidSAFE+ COPPA compliant and built to support healthy, active play for all ages and abilities. Nex Playground features a growing library of 50+ experiences, including motion and dance games, fitness and educational experiences, and Nex Originals. Content includes collaborations with partners like Hasbro, Sesame Workshop, and NBCUniversal. Nex has been recognized by Fast Company’s Most Innovative Companies, TIME’s Best Inventions, and Parents’ Best Entertainment System for Families, and has earned Red Dot, IDEA, and Core77 international design awards. We encourage you to explore Have Fun and Is Motion Gaming Back? , as they offer a deeper look into our culture, values, and explain how our approach to motion gaming differs from previous generations. Nex employees may be expected to collaborate with colleagues across different time zones. Depending on the role and location, occasional travel may also be required for company events, external meetings, and other business-related activities. Important Notice: ⚠️ To protect yourself from recruitment scams, please note that Nex communicates through official company emails ending in @nex.inc . If you receive a suspicious email or message, please do not respond and report it immediately to us. Location: Hong Kong Type: Full Time Internship (June 1, 2027 - August 31, 2027) The Role The Core Technology team develops the technologies that allow Nex Playground to use cameras, microphones, and other sensors to understand player movement and interactions in real time. As our Software Engineer Intern (ML Research, Core Technology) , you will explore and develop machine learning methods that expand what Nex Playground can perceive and understand. You may work on computer vision, audio, sensor fusion, or other sensing problems, taking ideas from research hypotheses through experimentation, evaluation, optimization, and working prototypes. You will collaborate with ML researchers and software engineers to build datasets and experimental workflows, evaluate ideas rigorously, and demonstrate promising technologies under real-world and on-device constraints. You will receive hands-on mentorship while owning a meaningful research project that contributes to our future products and technical direction. What You Will Do Develop and evaluate ML models and algorithms for computational perception and interaction. Design rigorous experiments, improve datasets, compare approaches, and understand failure modes. Build training and evaluation tools, and optimize promising models for real-time, on-device inference. Collaborate with researchers and engineers to prototype technologies on Nex Playground and communicate your findings. Should Have Must be currently enrolled in an undergraduate or graduate degree program at an accredited college or university. Strong interest in machine learning and enthusiasm for learning new methods and tools. Solid mathematical fundamentals and comfort with quantitative reasoning. Strong analytical and problem-solving skills. Programming experience in any language. A curious, systematic approach to learning through experimentation. Strong communication and collaboration skills. Nice To Have Experience with Python or familiarity with PyTorch, TensorFlow, or another ML framework. Interested in computer vision, audio processing, sensor fusion, or multimodal learning. Interested in model optimization, on-device inference, GPUs, NPUs, or resource-constrained devices. Interested in data curation workflows, or dataset quality analysis. Enthusiastic about games, interactive physical experiences, and helping people be active and healthy. We Offer Competitive compensation package. Flexible working hours and vacation policy. Product-driven culture that treasures talents and individual growth. Front-row seat and hands-on experience with cutting edge technologies in the evolving gaming field Nex is located in San Jose, California, USA and Hong Kong . Learn more about us at https://www.nexplayground.com/about . We encourage applications even if you don’t meet more than 50% of the requirements — we believe that experience comes in many forms! Nex is an equal opportunity employer and does not discriminate on the basis of race, color, religion, sex (including pregnancy, childbirth, gender identity, and sexual orientation), national origin, ancestry, age, physical or mental disability, medical condition, genetic information, marital status, military or veteran status, or any other characteristic protected by applicable law. This policy applies to all individuals at every stage of the employment relationship, including all current and prospective employees, and covers all employment decisions including recruitment, hiring, job assignment, promotion, compensation, benefits, training, discipline and termination. We are committed to providing a workplace free from discrimination and harassment and to fostering an inclusive environment for all employees.

Задачи: — Разработка и внедрение автономных ИИ-агентов и генеративных моделей (LLM, мультимодальные системы) для ускорения проектирования, разработки и развертывания ERP-систем, а также создания интеллектуальных интерфейсов автоматизации — от генерации кода до обработки естественно-языковых запросов пользователей. — Создание ИИ-сервисов и автономных агентов для автоматизации бизнес-процессов. — Обучение и тонкая настройка LLM (Large Language Models) и других ML-моделей под задачи компании. — Анализ данных, разработка алгоритмов для прогнозирования, оптимизации и принятия решений. — Интеграция ИИ-решений в существующие продукты компании. — Участие в научно-исследовательских проектах, публикация результатов (при желании). Требования: — Опыт в работы в IT от 5-ти лет. — Опыт работы в области Machine Learning, Deep Learning или Data Science от 3-ех лет. — Знание Python и ряда библиотек для ML/DL таких как TensorFlow, Pandas, PyTorch, scikit-learn и др. — Опыт работы с NLP, LLM (GPT, Llama, Mistral и т. д.), RAG, автономными агентами. — Понимание принципов работы ERP-систем (желательно). — Умение работать с большими данными (SQL, NoSQL). — Знание методов обучения с подкреплением (RL) — будет плюсом. — Английский язык (уровень для чтения технической документации). Условия: — Официальное оформление и стабильный доход по ТК РФ. — Все льготы и преимущества работы в аккредитованной ИТ-компании. — Программа ДМС с первых дней. — Работа в офисе с гибким началом рабочего дня. — Московский офис в шаговой доступности от м. Динамо, оборудованный всем необходимым для комфортной работы и отдыха.

Location: New York; Montréal; Toronto Workplace: Remote ABOUT MISTRAL Mistral provides full-stack AI solutions: from frontier models to developer tools, applications, and compute. We partner with enterprises tackling the hardest problems across high-stakes industries like finance, manufacturing, defense, healthcare, and the public sector, co-creating customized AI systems that they can run on their terms. We are a dynamic, collaborative team passionate about AI and its potential to transform society. Our diverse workforce thrives in competitive environments and is committed to driving innovation. Our teams are distributed between Europe, North America, Asia and the Middle East. We are creative, low-ego and team-spirited. THE ROLE As a Compute Solution Architect, you will support our strategic AI Infrastructure customers. In this role, you will be a trusted technical advisor, helping clients design, deploy, and scale AI solutions while managing large-scale GPU workloads involving thousands of GPUs. You will also collaborate with sales and product teams to drive growth and enhance customer satisfaction. WHAT YOU WILL DO Your responsibilities will include: - Serve as the primary technical point of contact, troubleshooting and resolving for AI Infrastructure services. - Lead the delivery of strategic GPUaaS contracts and ensure all commitments are met (timelines, SLAs, support, …). - Act as a bridge to product and support teams, providing customer feedback, relaying feature requests, and ensuring alignment with customer requirements. - Guide customers in optimizing GPU performance for ML training and inference workloads, ensuring seamless integration and scalability. - Partner with the sales team to identify new opportunities, promote the latest products, and deliver technical presentations. WHAT WE'RE LOOKING FOR - 5-10 + years in roles like Cloud Solutions Architect, Technical Account Manager, or Customer Engineer, with hands-on experience in cloud services and AI/ML workloads. - Track record as a technical pre-sales/solutions architect driving infrastructure engagements with partners through to production. - Strong understanding of AI/ML concepts and advanced implementations. - Proven success driving technical engagement and technical adoption. - Working knowledge of cluster orchestration and scheduling: Slurm, Kubernetes, and batch/HPC workload management. NICE TO HAVE - Working knowledge of model training frameworks runtimes (e.g., PyTorch, JAX, Tensorflow) and Generative AI serving runtimes (e.g., vLLM, TensorRT-LLM, SGLang, TGI, LLM-D). - Proficiency with parallel file systems (e.g., Lustre, GPFS) and object storage integration. - Familiarity with high-performance infrastructure: InfiniBand/RDMA networking, parallel storage, and multi-node distributed training topologies. WHY THIS ROLE MATTERS Compute deals require flawless execution from the initial technical commitment through deployment and production. By coordinating customers, sales, product, and support, the Compute Solution Architect ensures that infrastructure performs as expected, contractual commitments and SLAs are met, and issues are resolved quickly. This end-to-end ownership is critical to building customer trust, securing long-term adoption, and turning strategic GPUaaS contracts into successful partnerships. HIRING PROCESS - Talent Acquisition screening call - 45 min Hiring Manager & Business Acumen interview - 60 min Technical deep dive - Panel (home prep & restitution) - 30 min Value talk - Reference checks WHAT WE OFFER We offer a comprehensive benefits package designed to support your well-being, growth, and work-life balance. Benefits vary by country and may include healthcare coverage, parental leave, retirement plans, relocation support, wellness programs, meal and transportation allowances, and other location-specific perks. For the most up-to-date details on benefits available in your location, please refer to our Benefits page https://app.notion.com/p/mistralai/Benefits-at-Mistral-36e6ba59a7fe836b93dd01737fcc27ef?source=copy_link. PRIVACY POLICY Your privacy matters to us. You can learn more about how we handle your personal data in our Applicant Privacy Policy https://legal.mistral.ai/terms/applicant-privacy-policy.

Location: San Francisco, CA About Anthropic Anthropic’s mission is to create reliable, interpretable, and steerable AI systems. We want AI to be safe and beneficial for our users and for society as a whole. Our team is a quickly growing group of committed researchers, engineers, policy experts, and business leaders working together to build beneficial AI systems. About the role: The Safeguards ML Inference Path team designs, builds, and operates the production infrastructure that powers Claude's ML based safety systems. We collaborate closely with safety researchers and inference engineers to bring new classifiers and novel classes of ML defenses to production. We own the research → production transfer of new safety technologies that is on the critical path for every Claude model launch. And we build for scale: serving thousands of ML classifiers, for all requests on the token generation path, and for every platform Claude runs on -- 1P, Bedrock, Vertex, and beyond. We’re growing the team and looking for engineers who have deep expertise in productionizing ML systems. You'll work at the intersection of machine learning, large-scale distributed systems, and AI safety, developing the platforms and tools that enable our safeguards to operate reliably at scale. And your tooling and infrastructure will be used for every model launch, which are becoming more complex, and more frequent. Responsibilities: Design and build scalable ML infrastructure to support real-time safety deployments across our classifier and model ecosystem Build monitoring and observability tools to track classifier performance, data quality, and system health for safety-critical applications Collaborate with research teams to productionize safety research, translating experimental safety techniques into robust, scalable systems Optimize inference latency and throughput for real-time safety evaluations while maintaining high reliability standards Implement automated testing, deployment, and rollback systems for ML models in production safety applications Partner with Safeguards, Security, and Alignment teams to understand requirements and deliver infrastructure that meets safety and production needs Contribute to the development of internal tools and frameworks that accelerate safety research and deployment You may be a good fit if you: Are proficient in Python and have experience with ML frameworks like PyTorch, TensorFlow, or JAX Understand distributed systems principles and have built systems that handle high-throughput, low-latency workloads Have built automated or self-service deployment pipelines and eval infrastructure allowing researchers to roll out classifiers and models independently Have implemented A/B testing frameworks and experimentation infrastructure for ML systems Are results-oriented, with a bias towards reliability and impact in safety-critical systems Enjoy collaborating with researchers and translating cutting-edge research into production systems Care deeply about AI safety and the societal impacts of your work Strong candidates may also have experience with: Have 5+ years of experience building production ML infrastructure, ideally in safety-critical domains like fraud detection, content moderation, or risk assessment Working with large language models and modern transformer architectures Developing monitoring and alerting systems for ML model performance and data drift Experience in trust & safety, fraud prevention, or content moderation domains Knowledge of privacy-preserving ML techniques and compliance requirements The annual compensation range for this role is listed below. For sales roles, the range provided is the role’s On Target Earnings ("OTE") range, meaning that the range includes both the sales commissions/sales bonuses target and annual base salary for the role. Annual Salary: $320,000 — $485,000 USD Logistics Minimum education: Bachelor’s degree or an equivalent combination of education, training, and/or experience Required field of study: A field relevant to the role as demonstrated through coursework, training, or professional experience Minimum years of experience: Years of experience required will correlate with the internal job level requirements for the position Location-based hybrid policy: Currently, we expect all staff to be in one of our offices at least 25% of the time. However, some roles may require more time in our offices. Visa sponsorship: We do sponsor visas! However, we aren't able to successfully sponsor visas for every role and every candidate. But if we make you an offer, we will make every reasonable effort to get you a visa, and we retain an immigration lawyer to help with this. For roles performing work in California: We will consider for employment all qualified applicants, including those with arrest or conviction records or criminal histories, in a manner consistent with the requirements of applicable state and local laws, including the San Francisco Fair Chance Ordinance, the City of Los Angeles’ Fair Chance Initiative for Hiring Ordinance, the Los Angeles County Fair Chance Ordinance for Employers, and the California Fair Chance Act. We encourage you to apply even if you do not believe you meet every single qualification. Not all strong candidates will meet every single qualification as listed.  Research shows that people who identify as being from underrepresented groups are more prone to experiencing imposter syndrome and doubting the strength of their candidacy, so we urge you not to exclude yourself prematurely and to submit an application if you're interested in this work. We think AI systems like the ones we're building have enormous social and ethical implications. We think this makes representation even more important, and we strive to include a range of diverse perspectives on our team. Your safety matters to us. To protect yourself from potential scams, remember that Anthropic recruiters only contact you from @anthropic.com email addresses. In some cases, we may partner with vetted recruiting agencies who will identify themselves as working on behalf of Anthropic. Be cautious of emails from other domains. Legitimate Anthropic recruiters will never ask for money, fees, or banking information before your first day. If you're ever unsure about a communication, don't click any links—visit anthropic.com/careers directly for confirmed position openings. How we're different We believe that the highest-impact AI research will be big science. At Anthropic we work as a single cohesive team on just a few large-scale research efforts. And we value impact — advancing our long-term goals of steerable, trustworthy AI — rather than work on smaller and more specific puzzles. We view AI research as an empirical science, which has as much in common with physics and biology as with traditional efforts in computer science. We're an extremely collaborative group, and we host frequent research discussions to ensure that we are pursuing the highest-impact work at any given time. As such, we greatly value communication skills. The easiest way to understand our research directions is to read our recent research. This research continues many of the directions our team worked on prior to Anthropic, including: GPT-3, Circuit-Based Interpretability, Multimodal Neurons, Scaling Laws, AI & Compute, Concrete Problems in AI Safety, and Learning from Human Preferences. Come work with us! Anthropic is a public benefit corporation headquartered in San Francisco. We offer competitive compensation and benefits, optional equity donation matching, generous vacation and parental leave, flexible working hours, and a lovely office space in which to collaborate with colleagues. Guidance on Candidates' AI Usage: Learn about our policy for using AI in our application process.

Location: Europe We are toogeza, a Ukrainian recruiting company focused on hiring talent and building teams for tech startups worldwide. People make a difference in the big game, and we may help find the right ones. Currently, we are looking for a Machine Learning Engineer — Physics AI for Zibra AI. Zibra AI is a deep-tech company building advanced technologies for working with large-scale 3D data. The team has a strong background in computer graphics and data compression and is now expanding its technology into industrial simulation and Physics AI. The company is developing a new data infrastructure layer that makes massive scientific and simulation datasets significantly easier to store, transfer, visualize, and use for AI training. You will work at the intersection of Physics AI, scientific computing, ML systems, and data compression. A major part of the role is to benchmark our codec across different model architectures, study how compression affects accuracy and training efficiency, and explore new approaches to training directly in compressed representations. WHAT YOU WILL DO - Benchmark our compression technology across a wide range of Physics AI architectures and datasets. - Run large-scale experiments for CFD, turbulence, weather, engineering, and other scientific ML workloads. - Measure the impact of compression on: - model convergence and final accuracy; - training throughput; - GPU utilization; - CPU and data-loading overhead; - storage and network requirements. - Compare compressed-data training against conventional pipelines and alternative compression methods. - Research training directly in compressed or partially decoded representations. - Explore compression-aware sampling, augmentation, tokenization, and model architectures. - Design rigorous, reproducible benchmark methodology. - Integrate compressed datasets into PyTorch and distributed training workflows. - Turn experimental results into product recommendations and research directions. - Write technical reports, benchmark publications, blog posts, and academic papers. - Collaborate with external research groups and industrial partners on joint evaluations. WHAT WE ARE LOOKING FOR - Hands-on experience with Physics AI / scientific machine learning is required. - Experience training models on simulation or physical-science datasets. - Strong practical experience with PyTorch and modern deep-learning workflows. - Familiarity with architectures such as: - neural operators; - mesh GNNs; - transformers for physical systems; - surrogate models; - foundation models for science; - PINNs or related methods. - Experience with large 3D/4D datasets such as volumetric grids, meshes, point clouds, or spatiotemporal fields. - Good understanding of GPU training performance, data loaders, profiling, and distributed training. - Strong experimental methodology and ability to design controlled benchmarks. - Ability to analyze how numerical approximation and preprocessing affect model quality. - Strong Python and scientific-computing skills. - Ability to communicate research results clearly in written technical form. NICE TO HAVE - Experience with PhysicsNeMo or similar scientific ML frameworks. - Background in CFD, FEA, climate, turbulence, combustion, or computational physics. - Knowledge of lossy compression, quantization, numerical error analysis, or signal processing. - Multi-GPU or multi-node training experience. - Previous academic publications in ML, scientific computing, compression, or related fields. If this role sounds like a fit — we’d love to hear from you! Just send over your CV and anything else you’d like us to consider. We’ll review everything within five working days, and if your background matches what we’re looking for, we’ll get in touch to set up a call and get to know each other better.

Responsibilities: • Design multi-agent systems and integrate LLMs with internal tools • Develop backend APIs, RAG pipelines, and tool/function calling • Build computer vision pipelines for detection, tracking, and re-identification • Optimize real-time video and GPU inference Requirements: • Node.js/Express.js and Python • LLM APIs, prompt engineering, RAG, tool calling, or agent frameworks • Docker, Linux, and Git • PyTorch and OpenCV • Object detection and multi-object tracking (YOLO, ByteTrack/BoT-SORT) • Person re-identification, vector search, and video streaming (RTSP)