EmergingReasoning, planning and search

Reasoning Research Scientist

Also called Reasoning Scientist

A Reasoning Research Scientist tests and builds methods that make a model's multi-step inference reliable. They show the company which approaches give correct answers on hard problems and what those answers cost in compute.

What they do

Designs and runs experiments on how models reason through multi-step problems. Builds search environments where answers can be checked, then compares methods such as sampling many solutions, guided search and learned verifiers. Writes evaluations on verifiable tasks, so that correctness is not a matter of opinion. Uses symbolic solvers to check outputs. Measures how results change as inference compute rises or falls. Reads failed reasoning traces closely to find where chains break. Reports findings plainly to engineers who decide what ships.

What a strong hire delivers

In the first months, a strong hire delivers a trusted evaluation set of verifiable problems with held-out splits. They produce a baseline of current reasoning performance and a clear picture of how it moves with compute. They run a first comparison of two or three methods and recommend which to pursue. They also leave behind reusable search environments and written notes the team can build on.

How you can tell they are good

Ask them to walk through a reasoning method they tested and how they knew it worked. Strong people explain how they kept test problems held out, how they verified solutions, and what happened when compute changed. They can describe a result that failed and why. Be wary of anyone who cites benchmark gains without discussing contamination, variance or cost.

Why it is hard to fill

The title is rare and the work often sits inside other jobs, such as ML researcher, applied scientist or optimisation engineer. Few people have run reasoning evaluation end to end. Many candidates are strong in one strand, reinforcement learning or classical search, but not both. Job titles rarely show who has done the real work, so we read papers, code and project detail rather than CVs.

Where to find them

Look in research teams at frontier labs and well-funded AI start-ups, and in academic groups working on reasoning, planning or verification. Others come from operations research, algorithms and combinatorial optimisation, or from formal methods and solver work. Some are reinforcement learning researchers who moved toward inference-time methods. Titles vary: research scientist, research engineer, applied scientist, optimisation scientist.

Indicative pay

Not enough published evidence yet to give a fair range for this role, which is common for work this new. We benchmark it for you as part of a search.

Interview questions for a Reasoning Research Scientist

  1. Describe a reasoning method you tested. How did you decide it actually improved correctness?

    A strong answer names the task, the baseline, the held-out set and how solutions were verified. They mention variance and compute. Worry if they rely on a single benchmark number or cannot explain how they avoided leakage.

  2. How do you build an evaluation for multi-step reasoning when answers can be checked automatically?

    Look for verifiable task design, difficulty spread, held-out splits, and checks on the reasoning rather than only the final answer. Worry if they ignore solver errors, ambiguous problems or ways the model could guess.

4 more questions, with what a strong answer covers, are in the Role Kit.

Job description you can post

We are hiring a Reasoning Research Scientist to make our models more reliable at multi-step problems. You will test methods for reasoning and search, build evaluations that can be verified, and tell us what works and what it costs. We value demonstrated work over titles.

What you will do

  • Design and run experiments on multi-step reasoning, planning and inference-time search
  • Build search environments and verifiable-task evaluations with held-out problems
  • Compare methods such as sampling, guided search and learned verifiers
  • Measure how correctness changes with inference compute
  • Analyse failed reasoning traces and report root causes
  • Write clear recommendations for engineering and research colleagues

What you will bring

  • Shown work testing reasoning or search methods, with results you can explain
  • Experience building evaluations with verifiable answers
  • Strong Python and comfort with search environments
  • Working knowledge of transformer models and reinforcement learning
  • Background in algorithms, ML research or operations research
  • Clear written reporting of results and limits

Nice to have

  • Experience with symbolic solvers or formal verification
  • Published or open work on reasoning or planning
  • Experience with verifier or reward model training

Free to use and adapt. If it saves you time, a link back to this page is appreciated.

Skills that matter

  • Transformer modellingAdvancedImplement attention-based sequence or multimodal models
  • Reinforcement learningAdvancedOptimise policies from interaction and reward
  • Optimisation and operations researchWorkingChoose actions subject to objectives and constraints
  • Model evaluationWorkingMeasure capability and failure modes credibly

Other roles in reasoning, planning and search

Hiring nowPlanning Algorithms ResearcherA Planning Algorithms Researcher builds planners that choose sequences of actions under hard constraints. Companies hire them so agents and systems…EmergingInference-Time Compute ResearcherAn Inference-Time Compute Researcher decides how much thinking a model should spend on each problem. They raise accuracy on hard tasks while keeping…EmergingNeuro-Symbolic AI ResearcherA Neuro-Symbolic AI Researcher combines learned models with symbolic rules and solvers, so systems reason consistently where rules matter. That…

Related roles in other families

Hiring nowPost-Training Research EngineerPost-training and preference optimisationA Post-Training Research Engineer adapts pretrained models into ones that behave as the company needs. They turn preference data and reward signals…Hiring nowReward Modelling ResearcherPost-training and preference optimisationA reward modelling researcher builds models that learn what good output looks like from preference comparisons. They give post-training teams a…Hiring nowReinforcement Learning Research EngineerPost-training and preference optimisationA Reinforcement Learning Research Engineer improves how a model behaves by training it against interactive signals. They matter because reward gains…Hiring nowNatural Language Processing ScientistLanguage and multilingual intelligenceAn NLP Scientist builds and tests methods for language understanding and generation. They make sure a company's language models work across…

How we wrote this

Role definitions are our own editorial work, checked against occupational classifications, standards, research and employer pages. Titles vary by company, and this is a description of the work, not a job advert.

← Reasoning, planning and searchAll new job families

Start a search

Hiring a Reasoning Research Scientist?

Tell us the work. We find the people already doing it, whatever their job title says, and you meet them already interviewed for your role.