Design and implement model routers that select models or reasoning profiles based on quality, cost, latency, reliability, and related objectives.
Develop provider and protocol abstractions, versioned schemas, execution contracts, and portable agent skill and memory systems.
Create benchmark suites and evaluation protocols for models, routers, memories, skills, harnesses, and agent workflows.
Research retrieval, context selection and compaction, identity, provenance, distillation, self-improving harnesses, multi-agent learning, and automated skill creation.
Write robust research software, APIs, integration layers, distributed systems, and test infrastructure for reproducible experimentation.
Collaborate across research and engineering teams and communicate findings through technical reports, demonstrations, open-source releases, benchmarks, and publications.
Requirements
Profound understanding of machine learning, large language models, or statistical decision-making.
Deep expertise in at least one relevant area such as model routing, recommender systems, agent systems, retrieval and memory, model evaluation, distributed systems, or protocol and API design.
Experience building and evaluating modern language-model or agentic systems with tool use and multi-turn workflows.
Experience designing, executing, and analyzing machine-learning experiments with statistical rigor, including held-out testing, out-of-domain evaluation, uncertainty, and reproducibility.
Strong software-engineering and algorithm-design skills, including excellent Python proficiency and experience with APIs, data schemas, distributed services, testing, observability, code review, and CI/CD.
Ability to reason about security, privacy, provenance, permissions, failure modes, and user control in agent systems.
Strong communication and technical leadership skills across research and engineering disciplines.
Preferred experience with model routers, cascades, mixture-of-experts, recommenders, cost-aware inference, multiple model providers, agent harnesses, retrieval systems, vector search, knowledge graphs, benchmark suites, distillation, reinforcement learning, preference learning, reward modeling, or automated skill generation.
Preferred proficiency in TypeScript, Go, Rust, or another systems language in addition to Python.
Preferred experience with secure authentication, sandboxing, privacy-preserving telemetry, policy-enforced execution, distributed data-processing, model-training, or inference systems.
A PhD in Computer Science, Machine Learning, Artificial Intelligence, or a related technical field, or equivalent practical experience, is preferred.
A track record of impactful publications, open-source contributions, deployed AI systems, or delivered products and research prototypes is preferred.
Benefits
Competitive compensation.
Career growth and learning opportunities.
Flexibility and ownership.
Collaborative and innovative culture.
Opportunity to work on impactful AI projects.
International environment with talented teams.
Salary: Competitive
Nebius
Nebius develops reliable and user-friendly platforms that help organizations improve efficiency and achieve their business objectives.