Applied Scientist
What the work looks like, the stacks behind the title, what each level means and the training that helps, then how interviews run and how to prepare for each step.
What an applied scientist actually does
A scientist who turns research methods into models that reach real users, and proves their value with rigorous experiments.
An applied scientist designs and builds machine learning models for real products and business problems. They choose the method, often from recent research, adapt it to the data of the company and train the models, often with PyTorch and Hugging Face.
Rigor is at the heart of the role. Applied scientists design experiments, use statistics and causal inference to measure the real effect of a model, and explain their results to engineers, product managers and leaders.
They work closely with ML engineers who put the models in production, and some of them publish their work at scientific conferences.
- Training and evaluating a new model approach
- Designing an A/B test or an offline experiment
- Reading and discussing recent papers with the team
- Presenting results to product and engineering
- Writing code and documentation with ML engineers
Large tech companies such as Amazon and Microsoft use the title Applied Scientist for scientists who build models for their products. Other companies call the same role Data Scientist, Machine Learning Scientist or Research Scientist.
The job ad shows where the role sits: closer to research when it asks for publications, closer to product when it asks for models in production and measured impact.
The same title, very different stacks
Most applied science searches are defined first by the domain the models serve, then by the ground every applied scientist shares.
What each level means in practice
Titles vary from one company to the next. What changes from one level to the next is the scope you own and the decisions you take.
- Joins after a PhD or a master's on a defined problem
- Learns the data, the tools and the product
- Runs experiments with guidance from a senior scientist
- Owns a modeling problem from data to results
- Designs experiments and reads results with rigor
- Works directly with ML engineers and product teams
- Leads the science for a product area
- Chooses the methods and the experiments to run
- Mentors other scientists and reviews their work
- Sets the science strategy across teams
- Leads choices on methods and long-term research
- Shapes the work of several teams and their roadmaps
The paths into the role, and what really counts
Applied science is one of the roles where education weighs most. Companies still look first at models that reached users and at the rigor of your experiments.
- PhD in machine learning, statistics or computer scienceThe most common path, and often expected by large tech companies.
- Master's in machine learning or statisticsA common path, often with research experience during the degree.
- Data science firstData scientists with strong modeling skills also move into the role.
- Engineering school with a research trackA common path in France, often followed by a PhD or a research role.
Our view: for this role, a model that reached users and an experiment that proved its value often say more than any certificate.
A conversation with a senior tech recruiter
Your first interview is with one of our senior tech recruiters, who knows the role, the stacks and the levels in this guide. The conversation covers your own work through real examples, and the technical exercises come later, with the company. Before the call, we send you the six themes we will cover so you can prepare your examples.
- What you owned yourselfThe models, experiments or research projects you were responsible for, and your own part in the team's work.
- Your environment and its constraintsData quality, product needs, deadlines, compute budget or rules on data: what shaped your scientific choices.
- The decisions you madeWhich method, which data and which experiment, the options you compared and the reasons for your choice.
- The impact of your workWhat changed thanks to your work, with numbers when you have them: model metrics, A/B test results or business impact.
- How you work with other teamsYour work with ML engineers, product managers and leaders, and how you explain results to them.
- What you want nextThe kind of problems, domain and scope you are looking for in your next role.
- Models that reached users
- Rigor of experiments
- Results you can explain
The same ground for every candidate. For a given role, every candidate answers the same final questions, built with the hiring manager around what the role really needs.
Ready for the next steps. After our call, we tell you how the company runs its interviews, who you will meet at each step and what each step is meant to assess.
What the company will assess, step by step
Each company runs its own process, and we walk you through its steps before you start. Applied science processes often combine the steps below, and the depth expected at each step grows with your level.
- A coding exerciseLive coding in Python, on algorithms or on a small modeling task with data.
- An ML depth and breadth interviewQuestions on models, statistics, evaluation and the methods of your domain, often in depth on your own research.
- A research presentationA talk on one of your past projects or papers, followed by questions from scientists on your choices.
- An applied caseA business problem to turn into a modeling approach, with the data, the metrics and the experiments to run.
- ConfirmedYou solve the exercise cleanly, know the basics in depth and explain your methods in simple words.
- SeniorYou lead the discussion, design sound experiments and point out the limits of your results before you are asked.
- Staff and PrincipalYou run most of the interview as a peer, connect methods to the strategy of the company and show how you guide other scientists.
How to prepare as an applied scientist
Practical advice for each step of an applied science process, from the way you introduce yourself to the questions you ask at the end of each interview.
Start with the models and research you worked on, their impact and your part in them. Mention your methods in one sentence, then close with what you want next.
For each one, prepare the problem, the data, your method, the options you set aside, the result and what you would change today. Say "we" for the team and "I" for your own part.
Prepare a 20 to 30 minute talk on one project, with the problem, the method, the results and the limits, for scientists outside your domain.
Interviewers often go deep on statistics and evaluation. Review hypothesis tests, metrics, overfitting and the methods of your domain.
In an applied case, start with the business goal and the decision it supports, then choose the model and the experiment.
Rules on AI coding assistants differ from one company to the next, and sometimes from one step to the next. Ask before each step, and when they are allowed, explain what you ask the tool and how you check its output.
- Walk us through a model you built that reached users.
- How would you measure the real effect of a new model?
- Your A/B test shows a small gain. How confident are you?
- Explain your research to someone outside your field.
- How do scientists and ML engineers share the work?
- How much time goes to research and how much to product?
- Does the team publish, and how is it decided?
- How does the team run and review experiments?
- Which compute and data can a scientist use?
- How does the team use AI coding tools day to day?
Where applied scientists come from, and where they go
The roles people most often hold before this one, and the ones they most often move to.
Many applied scientists also come from data science, and some move to research scientist roles.