Earlier this week, I had the opportunity to chat with someone who is the director of data science for a well-known social app. I can't publicly share our discussion, but I do have a couple of take-aways that have been echoed in other conversations:
Beef up the Github. A strong portfolio on Github is extremely important for showcasing your data science skills. This is a great way to demonstrate your capabilities despite not having a degree or work experience in software engineering or CS.
Yes, bootcamps are not a big deal to everyone, and that's okay. Companies are different. Some companies enjoy partnerning with bootcamps to employ their graduates. Some companies don't have big enough data science teams to actively recruit from bootcamps. Some companies are doubtful that a person can magically transform into a good data scientist after 3 months- and they'll either scoff at you, or realize you have potential and take you under their wing. What's harder for companies to differ on though, is your skills. Again, make a solid Github profile.
Companies have different definitions and expectations of a data scientist. Team sizes vary depending on the company. KPMG has ~200 data scientists (out of ~24000 employees), Allstate has ~100 data scientists (out of ~12000 employees), and this last company I talked to has less than 5 data scientists (I won't reveal size). So of course, data scientist roles will come with different responsibilities (some responsibilities may bleed into that of a data engineer). It's important that the data scientist's job is clearly outlined and that the team is organized in a collaborative and supportive manner, especially for smaller teams.
Also- I'm really enjoying the Data Skeptic podcast. There's a ton of interesting episodes (Detecting Terrorists with Facial Recognition?, Data Science at E-Harmony), and I'm adding their blog to my never-ending list of things to read...