EmergingScientific machine learning and research computing

AI Scientific Workflow Engineer

Also called AI Science Workflow Engineer

An AI Scientific Workflow Engineer connects data, models and instruments into reproducible science pipelines. They make sure results can be traced and trusted, so research teams can build on them rather than redo them.

What they do

They take the scattered parts of a scientific project and join them into one workflow. That means pulling in instrument data, cleaning it, running simulations or learned models, and recording what happened at each step. They write scientific Python, run jobs on HPC clusters, and tidy notebooks into code others can rerun. They work with scientists to write down assumptions such as units, boundary conditions and noise, then check the pipeline respects them. They design validation runs, track uncertainty through each stage, and keep records so any result can be traced to its inputs.

What a strong hire delivers

In the first months, a strong hire turns one fragile, person-dependent analysis into a workflow that a colleague can rerun and get the same answer. They produce validation records showing what was tested and against what. They document the scientific assumptions built into the pipeline, flag where models are used outside their valid range, and leave the team with clearer handovers and fewer unexplained differences between runs.

How you can tell they are good

Ask them to walk through a pipeline they built, from raw data to final result. A strong candidate can say where each assumption lives, how they checked it, and what would break it. Ask what they did when two runs disagreed. They should talk about uncertainty, provenance and validation unprompted. Be wary of anyone who only describes model accuracy or tooling choices.

Why it is hard to fill

Hardly anyone holds this title yet. The work usually sits inside research engineer, computational scientist or ML engineer roles, so job titles tell you little. The mix is rare: real scientific judgement, solid software habits and enough deep learning to be useful. Candidates tend to be strong on one side and thin on the others. Many scientists have never worked to engineering standards, and many engineers miss the science.

Where to find them

Look in scientific computing groups, research software teams and computational labs in physics, chemistry, biology, materials and climate. Titles include research software engineer, computational scientist, applied mathematician and scientific ML engineer. They come from universities, national research facilities, instrument makers, drug and materials discovery companies, and AI companies working on science. Many started as domain scientists who became the person who fixed the pipeline.

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 an AI Scientific Workflow Engineer

  1. Describe a scientific pipeline you made reproducible. What was broken before, and how did you prove it was fixed?

    Look for specifics: pinned environments, recorded inputs, fixed seeds, comparison of outputs before and after. A strong answer defines what same result means for that science. Worry if reproducibility is just saying they used version control.

  2. How do you capture and test the scientific assumptions built into a workflow?

    Good answers cover working with domain scientists, writing assumptions down, and encoding checks for units, ranges and conservation laws. They name tests that fail loudly. Worry if assumptions stay in people's heads or in comments.

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

Job description you can post

We are building AI systems for scientific discovery and need someone to join our data, models and instruments into workflows we can trust. You will work closely with scientists and ML engineers to make results reproducible, traceable and scientifically sound.

What you will do

  • Build reproducible workflows linking instrument data, simulation and learned models
  • Record provenance so every result traces to its inputs and settings
  • Write down and test the scientific assumptions each pipeline relies on
  • Design validation runs and keep clear validation records
  • Track and report uncertainty through each stage
  • Run and tune jobs on HPC systems
  • Turn exploratory notebooks into maintainable, rerunnable code

What you will bring

  • Shipped scientific pipelines that others have rerun successfully
  • Strong scientific Python and good software habits
  • Working knowledge of deep learning applied to scientific problems
  • Practical experience with uncertainty quantification
  • Depth in a scientific discipline or applied mathematics
  • Experience running work on HPC systems
  • Clear communication with domain scientists

Nice to have

  • Experience with differentiable simulators
  • Experience designing experiments alongside lab teams
  • Contributions to research software used by others

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

Skills that matter

  • Scientific domain expertiseAdvancedApply disciplinary theory and experimental standards
  • Deep learningAdvancedBuild and diagnose trainable neural models
  • Experimental designWorkingChoose comparisons that identify credible effects
  • Uncertainty quantificationWorkingEstimate confidence, calibration and model uncertainty

Other roles in scientific machine learning and research computing

Hiring nowScientific Machine Learning ScientistA Scientific Machine Learning Scientist builds learned models that obey the physics, chemistry or biology of a problem. They matter because a model…Hiring nowAI Research Software EngineerAn AI Research Software Engineer turns scientific machine learning ideas into software that other people can run, trust and extend. They give a…EmergingAI Reproducibility SpecialistAn AI Reproducibility Specialist independently reruns AI-supported scientific claims and reports what holds and what does not. They protect the…

Related roles in other families

Hiring nowComputer Vision EngineerComputer vision and spatial perceptionA Computer Vision Engineer builds detection and recognition systems that work in a defined setting, and proves how well they work. Companies rely on…Hiring now3D Vision ResearcherComputer vision and spatial perceptionA 3D Vision Researcher recovers spatial structure from images and depth data, then proves it holds up. Companies building robotics, mapping, AR or…Hiring nowVideo Understanding ScientistComputer vision and spatial perceptionA Video Understanding Scientist builds models that recognise actions, events and the dependencies between them in video. Companies hire for this…EmergingOpen-World Perception ResearcherComputer vision and spatial perceptionAn Open-World Perception Researcher builds vision models that know when they are looking at something unfamiliar. That matters because a system that…

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.

← Scientific machine learning and research computingAll new job families

Start a search

Hiring an AI Scientific Workflow Engineer?

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.