The role
Tavily, part of Nebius, is the web access layer for AI agents. Over two million developers use us, and our customers include Fortune 500 companies running agents in production.
We build and run our own index. That means we own the hard problems directly: crawl coverage at scale, freshness, cost per query, and the quality of what comes back. This role owns that infrastructure as a product.
When an agent asks a question, the answer has to be relevant, fresh, fast, and affordable. Those four things pull against each other, and someone has to decide which one gives. That is you. You will set what good means for each use case we serve and the bar a change has to clear before it ships, working closely with our evaluation team, who own how we measure it. You will also shape where our index goes next: what it needs to cover, how fresh it needs to be, and what it has to be good at.
The most open part of the job is what comes next. Search is being rebuilt for a consumer that is not a person. Agents do not want one result page, they want to break a problem apart, run many retrievals, and decide what to do next. That changes what search infrastructure should look like, and part of your job is to have a view on where ours should go.
The hard part is deciding what good enough means. Perfect coverage, perfect freshness, and low cost cannot all be true at once, and the person in this seat is the one who has to make that call for each use case and defend it.
In your first year, we expect you to set a clear quality bar for our search products that the team builds against, have a real influence on where our index goes next, and bring us a point of view on where search infrastructure is heading.
This is a hands-on senior IC role with real ownership, reporting to the Director of Product. You will work day to day with our search engineering and research teams, with the evaluation team, and with customer success, who bring us quality feedback from real accounts every week. Our search engineers are deeply experienced in large-scale retrieval, and the product team is small, so you will argue your case directly with the people building it. That only works if you have built at this scale yourself.
Your responsibilities will include:
Set the quality bar and release criteria for search: what good means per use case, and what a change has to clear before it ships. Our evaluation team owns the measurement, and you work closely with them on what to measure and why.
Own the depth, latency, and cost tradeoffs across our search products, and make them explicit and defensible.
Shape the direction of our index: what it needs to cover, how fresh it needs to be, and what quality bar it has to meet.
Build a point of view on where search infrastructure is going as agents, rather than people, become the main consumer, and turn that into a roadmap.
Know how we compare to the alternatives customers evaluate, and drive the work that closes the gaps that matter.
Work with customer success and directly with customers to understand where our results fall short in real workflows, and turn that into quality priorities.
Partner with the data and vertical product owners on what our corpus needs to contain and how results should be shaped per industry.
We expect you to have:
5+ years of product management experience, with a meaningful portion spent on high-scale search or ranking systems: web search, feed ranking at a large social platform, or a large ad engine.
A background as an engineer, ML engineer, or researcher on a production system in one of those areas. This is a requirement rather than a preference. You will be making tradeoff calls alongside senior search engineers, and you need to have built something at that scale yourself.
Direct experience shipping a search or ranking change end to end, including how it was evaluated and what happened after it shipped.
A strong data foundation: you can design an experiment, read eval results critically, and tell a real quality improvement from noise.
Product judgment to match the technical depth: you can decide what to build and what to leave, and defend it to both engineers and customers.
Strong understanding of LLMs, AI agents, and how retrieval fits into agentic workflows.
Comfort with ambiguity, moving between strategy and execution without losing the thread.
The ability to influence and align cross-functional stakeholders without formal authority.
It will be an added bonus if you have:
Depth in information retrieval, NLP, or applied machine learning beyond what the role requires.
Experience with index tiering, sharding, freshness pipelines, or cost-per-query optimisation at scale.
Experience with hybrid retrieval, semantic and lexical search, re-ranking, or human evaluation programs.
Experience in API-first or developer-platform businesses running both self-serve and enterprise motions.
Experience in a high-growth technology company, especially post-acquisition or startup-within-a-larger-org environments.
Benefits & Perks:
Competitive compensation
Career growth and learning opportunities
Flexibility and ownership
Collaborative and innovative culture
Opportunity to work on impactful AI projects
International environment and talented teams
What's it like to work at Nebius:
Fast moving - Bold thinking - Constant growth - Meaningful impact - Trust and real ownership - Opportunity to shape the future of AI
Equal Opportunity Statement:
Nebius is an equal opportunity employer. We are committed to fostering an inclusive and diverse workplace and to providing equal employment opportunities in all aspects of employment. We do not discriminate on the basis of race, color, religion, sex (including pregnancy), national origin, ancestry, age, disability, genetic information, marital status, veteran status, sexual orientation, gender identity or expression, or any other characteristic protected by applicable law.
Applicants must be authorized to work in the country in which they apply and will be required to provide proof of employment eligibility as a condition of hire.
If you need accommodations during the application process, please let us know.
