Are you looking for advice on how to explain machine learning projects in interviews? Or maybe you are more interested in hearing about what interviewers are looking for when they listen to descriptions of machine learning projects. Well either way, you are in the right place!
In this article, we tell you everything you need to know about explaining machine learning projects in interviews. First, we describe what the average interviewer is looking for when they listen to descriptions of machine learning projects. After that, we discuss what aspects of your project you should focus on when you describe a machine learning project in an interview. Finally, we discuss aspects of your project that you should not focus on when you describe a machine learning project in an interview.
What are interviewers looking for when you describe a project?
What are interviewers looking for when they listen to descriptions of machine learning projects? While the exact details of what interviewers want to hear will vary from company to company, there are some common themes that most interviewers are interested in. Here are a few examples of questions that interviewers might have in mind when you describe a machine learning project in an interview.
- Can you communicate clearly about technical topics? The first thing that interviewers want to understand when you explain your machine learning project is whether you are able to explain highly technical projects to people who do not have as much context as you do. This will help them understand whether you would be able to communicate the objectives and results of your project to non-technical stakeholders and technical stakeholders who come from different backgrounds than you do.
- Do you have a history of delivering value to the business? The next thing that interviewers want to understand when you explain your machine learning project is whether you have a history of providing business value to the companies you have worked for. This will help them understand whether you will be likely to deliver value to their company.
- What was your role in the project? Interviewers will also want you to understand what role you played in the project. This will help the interviewer understand what role you will be able to play on their team. Did you contribute to a project that another team member was leading? Did you lead a single person project from start to finish? Did you lead a multi-person project that required you to delegate work to team members?
- What type of data science work do you do? Interviewers will also want to understand what kind of data science work your team does. Different data science teams at different companies focus on different types of tasks. Some teams write production code to ship machine learning models, whereas others create prototypes that are shipped to production by other teams. Other data science teams focus more on analytics and experimentation than machine learning models. Interviewers will want to understand whether the type of data science work you did at your previous company is similar to the data science work that they do on their team.
- How do you approach new challenges? Interviewers will also want to understand how you approach new challenges and how you go about solving ambitious problems. Do you create a simple solution first then iterate? Do you dive head first into a more complex solution? This will give them more insight into how you work and how you solve problems.
What to focus on when presenting a machine learning project
What aspects of your project should you focus on when presenting a machine learning project? Here are some examples of details you should focus on when presenting a machine learning project.
- Background and context. Before you present any details of your machine learning project, it is important that you provide sufficient background and context on the industry or domain that your project was grounded in. This will ensure that your interviewers are able to understand your project and the role that your project played in the broader organization. If there is any industry specific terminology that you will use as you describe your project, make sure that you define that terminology up front. Explaining the necessary background and context is one of the most important parts of your task because if your interviewer doesn’t not have enough context to understand what your project is and why it is important, then they will not understand anything else you say.
- The problem you were solving. Once you have provided sufficient context, you should clearly lay out the problem that you were aiming to solve when you started working on your project. You should include a description of how this problem ties back to high-level business goals to ensure that your interviewer understands why the problem needed to be solved. Were you removing a scale inhibitor that was preventing your company from growing? Were you automating a task that reduced the need for manual labor? Were you increasing conversion rates in an important funnel?
- Constraints that made the problem hard to solve. It is also a good idea to discuss any constraints that made the problem particularly difficult to solve. This includes both technical constraints and constraints that were related to people and stakeholders. Were there limits on the amount of computational resources you had access to? Did the model need to make fast inference so that it could be incorporated into a user facing product? Were you reliant on obtaining resourcing from an organization that did not think your project was a high priority?
- The team you worked with to solve the problem. Once you have laid out the problem you were trying to solve, you should provide more information on the team you were working with to solve the problem. Were you working in isolation on a solution? Were you working with other data science team members? Were you working on a large cross functional team with members from many different projects? You should provide information on who led the project and who contributed in other ways.
- A high level explanation of your solution. After you discuss the problem that you were trying to solve and the team you were working with, you should next provide a high level explanation of the solution you created. This should include a high-level explanation of the data that you used to train your model and the model that you used to solve your problem. You should also touch on how your solution was deployed. At this point, your aim should be to tell a clear and concise story that is simple and easy to follow. You should not include intricate details about the inner workings of your solutions. If your interviewer wants to hear more about the details of your project, they will ask you.
- The impact of your solution. Finally, you should provide information about the scale of the impact that your solution had. If you have not fully deployed your solution, but you have estimated the scale of the impact that the solution might have then you can share those numbers.
What not to focus on when presenting a machine learning project
- Intricate methodological details. While you should provide a high level explanation of the solution that you created for your project, you should not include complex technical details about your project. You should keep your explanations high level enough that they can be understood by someone who is in another field. This is particularly true if your interviewer does not have context about the industry you work in or the type of model you used. While you should not offer up these details on your own, you should be prepared to answer questions related to these details if your interviewer probes into them.
- Questions to ask at the end of data science interviews
- How to explain machine learning projects in resumes