You are at the stage of hiring your first or next round of data science talent? Talent that exactly fits your domain, learns quickly, and adapts to your working environment? After working with more than 3,500 data scientists and seeing dozens of startups hiring, here are our learnings and mistakes to avoid in order to identify your ideal candidate(s) quicker.
At the end of this blog, you find a unique model we developed to hire the best performers for your team.
“Many startups follow traditional hiring practices for data science talent — but they shouldn’t. The brutal truth is that standard hiring and interviewing questions are fatally flawed.”
Avoid these common mistakes
Ask candidates about their prior experience, and you’ll discover how well they can articulate what has happened in their previous jobs. Ask them technical questions, and you’ll uncover their ability to repeat knowledge. Make them solve a ‘toy’ problem, and you’ll discover how quickly they solve toy problems.
There are many candidates who master the interview process but are actually completely ineffective data scientists in practice. Essentially, you want to see how a candidate can perform in the real world, and solve an actual problem that is relevant to your company, including their ability to handle messy data, their soft skills, and how quickly they integrate into your team.
Identify your ideal candidate
To make this happen, you must first have a clear understanding of how you want candidates to perform data science. At the highest level, you should be clear on the prototype or product your team will produce. Will it be primarily visualizations and analyses? Prototypes that are given to developers? Or applications that can be scaled and deployed in production environments?
Next, you should have a clear understanding of what you want successful candidates to do. Identify problems you would love to see a data scientist tackle. Knowing answers to how your team performs data science and what challenges you most want candidates to be able to handle, you can design a hiring process that closely reflects your working conditions. This means you should put candidates into an environment that closely resembles what their ‘day-to-day’ would be. If they can succeed in that environment during the interview process, then their chances of succeeding long-term are much greater.
Set up a real-world interview process
Now, setting up such an interview process needs lots of time and trial and error in order to succeed. Something that most startups who are looking for immediate success, can´t build quickly enough.
Good news is, at Omdena we have built one of the world’s largest data science impact case study databases, building AI and accelerating the growth of dozens of award-winning startups from around the world. As a result, we developed a new type of hiring process that can help you succeed in becoming an AI-enabled startup in a much faster and more reliable way.
Omdena´s Real-World Approved Hiring Process
Imagine you define an actual problem that you need to build a prototype or even a deployable solution. Next, you get access to a team of 50 pre-vetted engineers who will solve the challenge for you within eight weeks, and at the end you not only get the solution but also you can hire the top performers.
In other words, you go through an entire data science project from problem scoping, to data collection, to AI modeling, while filtering out the best talent for your startup.
This is what we call the Omdena real-world approved hiring process, which we have developed after working with 70+ award-winning impact startups and organizations.
For example, Estonian startup Wildlife Species OU built a forecasting tool for endangered wildlife species by running a two-month Omdena AI Challenge as part of our startup incubation program.
While building sophisticated models, which helped “to see what is possible while finding the right direction to invest their resources in”, they also hired a collaborator from the challenge to enhance their team (30 seconds video testimonial below)
Overcoming data challenges and hiring an ML engineer
One of the most common data science problems arises during the data collection and annotation part. Here is how we solved this for Solar AI, a Singapore-based startup incubated as a part of ENGIE Factory.
Solar AI ran an AI challenge with us to build computer vision models that can help to hyper-scale the deployment of distributed solar panels. However, even the most technically advanced algorithms cannot address or solve a problem without the right data. Having access to data is quite valuable, but having access to data with a learnable structure is the biggest competitive advantage nowadays.
Project Snapshot: CV model for rooftop detection
Our team of collaborators annotated thousands of rooftops, which resulted in a perfectly labeled dataset for building the models. At the end of the project, Solar AI hired a machine learning engineer from the project to join their team.
“Working with Omdena has been truly impactful. The partnership has enabled us to team up with diverse collaborators to test and iterate on new models at speeds we would not have imagined doing ourselves.”
If you are interested to learn more about how we help impact-driven startups to build AI solutions and grow, check out our programs below.
Hiring the perfect candidates
Related article: How to Build a Lean AI Startup
