HOW WE HIRE AT SNIPS

Rand Hindi
Jul 15, 2015 · 8 min read

When we think about how to build our team, the first thing we think about is the culture we want to create. At Snips, we don’t see ourselves as a company with employees that have jobs, but rather as a community of passionate people. And together, we are working on our long term vision to fundamentally change the way we use technology in our everyday life. As such, our recruiting process puts a lot of emphasis on passion, team fit, creativity & hacker spirit.

I won’t be covering how we built our talent dealflow. Rather, I will assume that there are already hundreds of applicants in the pipeline, all of whom seem to have the right qualifications. The question thus becomes: how do we actually pick the right ones for our team? This is all the more important in small organizations, since each new team member will have a massive impact on culture and performance.

Moreover, talking to so many applicants while being small can become very distracting and time consuming. The trick here is to filter candidates as early as possible so that we can concentrate our efforts on the very best people and give them the attention they deserve.

Our process is short but hard. It is designed to assess particular variables that we identified as being important to the team. Roughly speaking, each of them can be thought of as a score between 0 (no relevant skills) and 5 (amazing skills). This enables us to quickly determine where the candidate shines, and where they can improve. Eventually, we want to use this method to determine in which team they would fit best, with the objective of maximizing the total area of the team itself.

A good candidate with solid skills and still room to progress

It is important to note that this is NOT how we determine the level and package for someone (we will write a post about this later). Rather, this is a binary outcome to determine if someone should be hired, regardless of their current experience.

The variables we assess are (still a work in progress):

  • CULTURE FIT: How well does the candidate fit in the culture of the company? Do they understand us? Do they pass the Sunday test? Will they make friends with other team members? Do they share our values?
  • SKILLS: How good is the candidate at their current job? If they are a data scientist, how much do they actually know about it? How much experience do they have?
  • HACKER SPIRIT: Do they know how to get stuff done? Do they know when to cut corners? Are they creative in solving problems?
  • TEAM SPIRIT: Are they a team player? Will they help other team members even if there is nothing to gain? Will they do whatever is necessary to make the team move forward? Do they have empathy?
  • LEADERSHIP: Is the candidate capable of leading a team? Are they recognized in their field? Will they attract other people to the company? Will they step forward during a crisis without taking pride?
  • MOTIVATION: How much does someone want to solve the problem we are tackling? Do they really want to join our team? Why? Will they put in the effort to make sure we solve the problem?
  • ROOM TO PROGRESS: Does the candidate still have room to progress? or are they already at their max?
  • CURIOSITY / PASSION: Do they genuinely care about what they do? In their free time, what do they do? Were they interested in it before it became fancy?

The 5-step process we designed to make this assessment is described below.

Step 1. The Application Form

Initially, we were asking people for their CVs. What we found was that this was not representative of how good they were, as it was really boring to read and emphasized things that we didn’t necessarily care about, while missing some key aspects. For instance, they typically do not include much indicators of creativity, team spirit and motivation to join the company, which are all determining factors for us.

So we ditched CVs completely and instead we now ask to fill a simple application form (check it out here). It starts with simple questions such such as name, email, Github, etc.., followed by bullet points about typical resume stuff, where each degree, experience or award is summarized in a single line (e.g. “PhD Bioinformatics @ UCL | 2011”). If we want to know more, we just ask about it later during a face to face conversation.

What we are actually really interested in are the answers we get to the other questions. For instance, we ask candidates to tell us about something they hacked. It doesn’t have to be software, it can be anything. What we look for here is a demonstration of creativity in getting stuff done with very little resources.

We then ask about a project that they loved, and what challenges they had to solve on the way. Here, we look for a demonstration of creativity, resilience, leadership and empathy. We are particularly interested in understanding how they potentially worked as a team to solve complex problems.

We also ask about what they would like to achieve in the next 12 months, even if it’s not related to the position they are applying for. What matters is that the candidate has a willingness to grow by learning and trying new stuff, which is what we do daily in our company. Curiosity, autonomy & ability to learn is very important when working on something no one has cracked yet!

Finally, we ask them about their motivation to join the team. We found that cover letters we received were very generic, and those who stood out had done their research on us and made up their mind before applying.

Step 2. The Challenge

This is my personal favorite. Once we selected candidates based on their application, we then send them a challenge based on the position they are applying for. For example, someone applying for a data science position will receive a real dataset and be asked to do something creative with it. They can use any tool, language and library they want, and supplement the data with any other sources. The goal for us is not to obtain a solution to the problem, but rather to see how creative people are in a real-life scenario.

The challenge itself is hard enough that it takes several hours to complete, but easy enough that it doesn’t require a priori knowledge of the problem at hand. Most importantly, it assesses a candidate’s passion for the subject.

Once we receive the challenge, we take a thorough look at it, and assess the following:

  • creativity in attempting to solve the problem
  • expertise in applying the right algorithms
  • ability to clearly explain the approach taken
  • quality of the code written

By doing this challenge we found that only people that are truly passionate about what they do will take the time to complete the task. They will put extra efforts in nailing smaller details because they enjoy tackling the problem. Since we designed the challenges to be as close as possible to the actual stuff we face in our daily work, someone passionate about solving our challenge will most likely be a very good fit.

Many applicants end up dropping out of the process at that stage, which is good, since we can then focus on the really passionate ones ;-)

Step 3. The Technical Interview

If we like the challenge and have a good feel for a candidate, we invite them to come meet our team and have a technical interview. The goal here is to dig deep into the knowledge of the candidates, and see what they would need to learn before they can become operational.

We know whiteboard interviews are not always the best way to assess the fit for engineers, so we never ask things like coding the A* algorithm as if it were a computer science exam — although we value deep algorithmic & theoretical computer-science knowledge.

What we do instead is try have a very deep discussion about a technical problem we are facing as a company. It helps us assess:

  • the technical depth of the candidate
  • human skills, communication skills and cultural fit
  • creativity, thinking speed, and pragmatism of solution
  • ability to pick up new concepts

Successful candidates often provide very useful insights into the problems we are having. They enjoy the discussion, and it generally feels more like an extremely interesting and deep discussion rather than a technical interview.

Step 4. The Hacking Trial

Based on the previous steps, the team determines whether the candidate would be a good fit, in which case they get invited for a trial period at our Paris office. The trial lasts 2 days, during which we must ship something together.

This trial period aims at testing both how productive the person is when working on hard problems, as well as if they are a good fit for the company and the current team. This of course works both ways, and gives candidates a good idea of what their life at Snips would be like.

Culture fit is in fact so important that we now have a simple rule: when there is a doubt, there is no doubt. To put it simply, you cannot “become” a Snipster. Either you are one naturally and the team embraces you, or you aren’t.

Step 5. The Reference Check

If the team likes the candidate, we ask for at least 2 references, which must be people the candidate has worked with. Although we didn’t do that in the beginning (it’s kind of cumbersome honestly), we have now started being very diligent about it, as we found it to be a great source of insights on a candidate.

During the reference call, we go deep and try to understand how the candidate solves problem, works in a team, what their qualities and shortcomings are (this requires creativity since references will typically only try to be positive towards the candidate), etc. What we want to find out are the hidden qualities or red flags.

Finally, We Celebrate!

We know our process can be tedious and hard. We built it that way because we want to make sure everyone joining the company does it for the right reasons. The worse things would be for someone to leave after few months because they (or we) realized it was not a good fit.

In fact, our process works so well that new team members barely feel “new” when we make the actual offer. It is a minor formality that, although duly celebrated, is overshadowed by the participant’s eagerness to pursue their work with us on game-changing problems, with a team so passionate that nothing else matters. As a founder, I feel humbled that so many exceptional people joined our quest for solving everyday problems using AI.

Executing on an very ambitious vision requires a great team, and as such, we believe that how we hire is not just important, it is actually the key to our long term success!

Snips is hiring!

If you care about creating products that will change the way we use our devices in our daily lives, take a look at our jobs page! We would love to hear about what makes you tick, your own personal projects, and discuss how we could work together!

Snips Blog

This publication features the articles written by the Snips team, fellows, and friends. Snips started as an AI lab in 2013, and now builds Private-by-Design, decentralized, open source voice assistants.

Thanks to Yann Lechelle

    Rand Hindi

    Written by

    #AI and #Blockchain entrepreneur working on Making Technology Disappear. I love to geek out on #Privacy! CEO @ snips.ai

    Snips Blog

    This publication features the articles written by the Snips team, fellows, and friends. Snips started as an AI lab in 2013, and now builds Private-by-Design, decentralized, open source voice assistants.

    Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
    Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
    Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade