Staff Applied Scientist Document Vision

Posting Type Remote/Hybrid Job Overview The Work Every legal matter is its own experiment. An attorney arrives with a theory of the case; the evidence arrives as hundreds of thousands of documents, sometimes millions, that no one has read and no model has seen. Somewhere in the cross product of the two are the answers that decide lawsuits, investigations, and livelihoods. Finding them quickly and defensibly, with the integrity and credibility attorneys can rely on, is the problem we own. We solve it creatively and rigorously. Relativity is a data-centered, AI-native legal technology company, and Applied Science builds the AI inside Relativity aiR. We launched aiR in 2023 and have now run commercial generative AI in the legal domain for more than three years, powering work that includes the largest investigations in the world. Our systems are distinguished by the data they operate over (more than 93 petabytes) and the work they have done: over 190 million AI review decisions, backed by more than 1 billion generative sub-analyses in 2026 alone. The team is as distinctive as the data: legal experts, all former litigators, work directly inside Applied Science. At Relativity, our mission is to Organize data. Discover the truth. Act on it. The Applied Science team serves this mission by building bold and ambitious AI systems. We are curious, dedicated, and humble. We understand complexity, uphold rigor, and measure relentlessly. We build and ship with pace. Above all, we are interdisciplinary collaborators and team players. We're looking for a Staff Applied Scientist to take on our hardest problems in document vision and set standards that reach beyond a single team. Job Description and Requirements Capable and Reliable Two requests can look nearly identical and be worlds apart. "See if you can find me an example of this" needs a capable system: it finds the example or it doesn't. "Conduct a reasonable search for any and all documents responsive to this request" is a different kind of promise. Its answer spans a corpus no one will ever read end-to-end. So the system's process, as much as its output, has to earn the trust of the professionals who rely on it. That property is reliability. It decomposes into consistency, robustness, calibration, and safety: systems that behave tomorrow the way they did today, degrade predictably under stress, know how confident they should be, and check their own work. Before aiR returns an analysis, it validates its citations and runs internal consistency checks; when a check fails, it refuses to answer. It has refused more than a million times so far in 2026, and we count every one as a success: an error caught before it reached a user. You'll build for both, and help define the standard for how. The Focus: Document Vision This role anchors our document-vision work: teaching systems to read evidence the way legal professionals do. Real matters arrive as scanned pages, photographs, tables, handwriting, stamps, and broken layouts, at the scale of millions of documents. You'll own the science of multimodal document understanding across aiR, from vision-language modeling to the evaluation standards that make visual evidence usable and defensible. What You'll Do Take on the hardest, most ambiguous problems in the portfolio and produce clarity: a well-specified approach, an evaluation that settles the question, a system that ships. Set standards that reach beyond your team: evaluation methods, modeling patterns, and quality bars adopted by scientists you've never worked with. Own readiness for flagship AI systems, from problem framing through efficacy studies and production monitoring, in partnership with engineering. Multiply the team through deep review and mentorship of senior and lead scientists. Advise Applied Science leadership on where the science is going and where we should invest. What You Bring A master's or PhD in computer science or another quantitative discipline (or equivalent professional experience), and at least 6 years of applied AI/ML experience. Years of production AI/ML behind you: systems you specified, shipped, and operated at scale, in partnership with the engineers who run them. Expert judgment about AI systems: you've built them, measured them, and formed views of their limits that hold up under challenge. Scientific rigor other people borrow: you supervise the data understanding of teams beyond your own, your evaluations become the template, and your error analyses end debates. Software-engineering judgment trusted across teams, and the programming skill to credibly prototype what you propose. An ownership mindset that extends across the organization. Nice to Have Experience with vision-language or document-understanding models (layout analysis, OCR-adjacent pipelines, table and figure extraction) in production. An interest in legal technology and the justice system. Experience developing information retrieval systems. Experience developing agentic harnesses. Experience building reliable AI systems at scale. Why Relativity Applied Science This is the place where your curiosity, dedication, and talent will build products that power the pursuit of justice around the world. Relativity is committed to competitive, fair, and equitable compensation practices. This position is eligible for total compensation which includes a competitive base salary, an annual performance bonus, and long-term incentives. The expected salary range for this role is between following values: $197,000 and $295,000 The final offered salary will be based on several factors, including but not limited to the candidate's depth of experience, skill set, qualifications, and internal pay equity. Hiring at the top end of the range would not be typical, to allow for future meaningful salary growth in this position. Required Skills: Algorithms, Computer Vision, Data Analysis, Data Science, Deep Learning, Machine Learning (ML), Natural Language, Natural Language Processing (NLP), Python (Programming Language), Scientific Research