Machine Learning opportunities can be sparse, so when you finally get invited to that long-awaited Machine Learning Interview, you want it to go perfectly. Let me show you how.
There is no go-to source for Machine Learning Jobs and many are distributed amongst general job boards, which include:
Arguably, writing your application and structuring your CV is a whole other post, so we'll leave this out for now. Reach out to all opportunities that interest you, don't wait for feedback on one, before moving to the next, and soon enough, you'll get to the next part.
It's time! It has been a few days since you applied, but a hiring manager has finally reached out to you to talk about the next steps. This is already the first opportunity to make a good first impression.
- Answer within an appropriate timespan (<24h)
- Give the hiring manager a selection of times and dates that work for you
- Subtly tell them that you are happy/excited about the opportunity
These might be obvious to you, but these things already hint how excited you are about the opportunity, if you are an organized person and how you communicate. With all these pieces of advice, keep one thing in mind: Stay subtle. There is no reason to reply a few seconds after the manager has sent a mail, neither is their one to use more than one '!' after your sentences.
As already described, the first interview is mostly a way to see if there is any fit between you and the company. In some cases, the very first interview might already be a technical screen, however, this is uncommon.
In this interview, you'll have to pitch yourself. Make it as easy as possible for the hiring manager to discover your talents. Research what the company does and where you can fit in. In some cases, a company might ask you 'What makes you excited about our company?'. If you don't know their products, this might be a TKO.
- Research what the company does
- See who their team members are
- Find out for yourself where you might fit in
- See if there is any match between your skills and their needs
- Prepare questions for things that you actually want to know
In many cases, this will be a phone/video screen, so make sure you are in a quiet space, with a stable internet connection and a good mic. There is nothing more irritating for the hiring manager than a buffering video stream, echo or background noise. And then the interview starts.
At this point, just be yourself. You have prepared yourself for the technicalities, but the rest is all you. Many companies will want to get a feeling for your character, personality, and even humor. Be calm and relaxed and if the situation is right, hint towards the things that you have prepared. Don't go against the natural flow of the conversation to pitch yourself, since that sounds extremely staged and dishonest.
In most cases, you will also have time to ask questions. Be sure to prepare these beforehand and only ask questions that you want to know, don't prepare filler questions that you don't care about. It's no shame to say that you don't have any questions. Good questions for Machine Learning might include:
- How many team members are there?
- What hardware do you use?
- Are you also doing data labeling?
- Do ML Engineers have to integrate their models somewhere?
After the interview has ended, it is sometimes a nice surprise to follow up with an email (maybe a day later) to thank them for their time and tell them if/that you enjoyed the interview. It also reminds them to follow up with you, maybe with more interviews.
Most companies have a hiring process that includes two, three or more interviews (hint: Having only one interview is a huge red flag). These interviews will most likely be technical, so if you know what the company does (e.g. computer vision) and you're a bit rusty in that area, be sure to brush up your knowledge in these areas.
In many cases, if you pass the first interview and you have a good skill set in Machine Learning, you shouldn't have to worry. Sometimes it works out, sometimes it doesn't. To test if you have a good skillset, you could try to go through some basic ML interview questions. If you get them all (or at least most) right, go out there and see what you can do!
The article was originally published here