Last week, Data Science student Anne-Sophie Hannes interviewed me about my own bootcamp experience (at another school), and how it changed my career. We also talked about my daily life as a co-founder & CTO at Ester.
A full recording of the talk is available. Registration is required.
No time to listen? Here is the gist of our conversation in a few paragraphs.
Cook up your own career
Combining two seemingly distinct areas of interest (e.g. natural language and computing) can go a long way in terms of crafting a career path. But when you’re stuck in a (real or perceived) dead-end, don’t hesitate to take a sidestep. That is, if you’re willing to take the risk. At the very least, it will bring you somewhere else.
Show, don’t tell
Whether you - or others - call you(rself) a web developer, data analyst, data scientist, ML engineer or AI engineer doesn’t matter all that much. Neither does trying to tick off all 27 boxes on an idealistic job profile. What does matter is a demonstrable project portfolio, and a passion for building stuff that works. Companies don’t hire profiles, they hire people. Make them want you.
Vectorize your beverages
The flavor experience of wine or beer is not only about the actual beverage, but also about consumers’ hyper-personal biological receptors, the shape and color of the container that holds the drink, the circumstances in which the drinks are consumed and many other contextual variables. Ester learns product representations from data sources that are as comprehensive as possible, in order to provide personalized recommendations.
Never waste a good crisis
In this world, unpredictable things happen. If it wasn’t for Covid-19, Ester would probably have had another CTO. The worldwide crisis that wreaks havoc, also opens opportunities for fast-movers that are able and willing to adapt. In the food & beverage industry, the strong push towards e-commerce is forcing incumbent retailers to accelerate their digital transformation strategies and programs. As an independent provider of flavor intelligence, Ester help retailers provide a better service to their customers, each time they figure out which products to like, choose and buy.
Organize your chaos
Startups tend to be inherently chaotic during the iterative search for the first minimum viable product (MVP) that has mimimal algorithmic performance (MAP). Precisely for that reason, it is vital to automate Continuous Integration and Deployment (CI/CD) processes early enough in the game. That is the only way to keep enough of the scarce resources available for actual value-added development work. Just as crucial is to pick the right cloud platform abstraction layer to develop on. If too high-level, there’s not enough flexibility. If too low-level, there’s too much infrastructure to manage. At Ester, we have chosen to be Kubernetes-neutral during the startup stage. That means Docker images and containers (and the code they run) are the lowest infrastructural elements we want to handle ourselves. Everything else is outsourced.
Follow fast, and say no - no - no
When making technology choices, it pays to be a little conservative when going from nowhere to somewhere in a matter of weeks. Suboptimal solutions that are good enough but delivered in time are much better (at first) than using the latest and greatest hyped yet risky technology. Saying no to that newest cool technology (GPT-3, anyone?) is hard, but right. Dare to go against the stream!
Could you smell that for me, please?
Digitized senses (aka artificial taste and smell) are an active area of research and development. Companies like Yesse Technologies are creating the building blocks for artificial mouths and noses that could augment the current sensory analyses conducted by human tasters in scientific labs like prof. Kevin Vertrepen’s at KULeuven/VIB.
Learn, steady, go!
Last, but not least: surviving a bootcamp was already hard enough in normal pre-Covid times. Don’t stress out too much about subjects that don’t fit you, instead focus on developing your strengths. Use your gut instinct, make choices and go for it. Don’t waste too much time trying to plan how to become successful. Time only runs forward!