Read on Twitter
So, the time has come to find somewhere to work as again. Some sort of Data Scientist / ML Engineer / Applied Scientist role. I have >10 years of experience and a relevant PhD. Let& #39;s see if any of you can see any of these statements in your picture:
You are a company with a data-centric product. I& #39;m looking for somewhere where I can apply some leverage on that data. Not somewhere where it is an exhaust of your core business and a cost center.
Quantitative product development: I design and perform experiments that move you forward. Data science is not about a big dashboard, but whatever artifact maximally reduces your uncertainty, clearing the fog of war. (I am still happy to work on your dashboard though.)
Picking the right goals: surveying, identifying and prioritizing potential things to solve. Not every problem with an upside is solvable, and not every solvable problem is worth pursuing.
Observability: running models live brings all sorts of fun failure modes. Every models should come with a monitoring system that& #39;s able to describe the envelope and scream on any anomalies.
Applied science: working on understanding what the core of your product is, and what scientific discpline has the right tooling to make it shine. Not everything fits in a LSTM (although maybe if you push hard enough, it will).
Maturation of models: iteratively refining and building a predictive model of your problem, refining as needed. Not about big models with long lead times. Having data scientists should reduce your stress, not increase it.
Multiplying the team capabilities: I don& #39;t like being the special wizard that is the only person that knows how to do and run something. My end game is that there is at least someone else that can play with my toys.
Proof-of-concepts: often, proving a hypotheses means building something. They can have very different lives. Some are just a Jupyter notebook, a LaTex PDF, and a faint memory, while some get shipped to customers as production ready. Neither are surprising.
Owning the data pipeline: having a good handle on how the data flows can make big differences on how well you can capture and model it. Tricky details happen in observation, collection, and moving it around. Happy to work on that, but also happy to let data engineers handle it.
Technology stack: the usual Python data toys (numpy, scipy, scikit-learn, pandas). Also familiar with Julia. Also Java/Scala, ideally Spark/Flink. If fast code needed, I can do C++ or extremely ugly python (numba, cython). If you& #39;re on the cloud, AWS is where I& #39;m most fluent.
Bonus: ideal job https://twitter.com/tadejtadej/status/1301475848835469315">https://twitter.com/tadejtade...
I& #39;m based in Berlin, so ideally local opportunities. Not moving now. Remote also considered (preferably Germany-based). Brief page here: https://tdj.si/ .">https://tdj.si/">... DM for CV.
