NBCUniversal - Research Assistant, Data Science & Predictive Analytics. x2
We're no startup. We're the special operations team for the corporate world.
We're hunting for engaging problems and opportunities to create real value for our marketing, sales, and research teams throughout our international businesses (television, movies, news, and home entertainment).
You'll be helping us on a variety of problems - little ones like writing a script to hack a third party's website and collect info on our products, to large ones like creating a model that can show our products value through time. In all cases you'll be working closely with the a senior engineer and either senior corporate management or a business unit in TV, film, or news. A typical week might look like this:
* Meet with the head of sales for a UK TV channel, learn about a problem they're facing trying to get an accurate value for advertisement slots.
* Discuss what you learned with the rest of team. Someone suggests that a nearest-neighbors regression on data from the marketing team might help.
* A senior dev helps you implement the algorithm and a UI for the sales team to use it.
* Spend some time with the sales team to see if it's creating any value, look for improvements.
What we use (We list this to tell you what we're like, not to tell you what languages we expect to see on a CV):
* Racket. We're big fans lisp and functional programming.
* Bash. We create a lot of value with small dirty hacky shell scripts, and we aren't shy about it.
* Whatever gets the job done, and whatever we think might be fun to use. Python, R, Javascript, Hadoop, EC2. We even wrote code in Fortran once (we don't like to talk about it.)
Who we're looking for:
Most importantly, you should be an intelligent problem solver who wants to apply whatever programming and maths knowledge you have to real world problems. We're not fussed about what languages are on your CV.
You have some familiarity with functional programming, machine learning, and linux/bash scripting, and you're eager for a chance to use it on real-world problems with support from other programmers. Given the predictive modeling angle, you're probably studying or have studied maths, stats, or even physics.
These are two assistant/internship positions for students looking for experience while doing their degree. Alternatively, if you can take on several projects in a full time role, please get in touch, too. We're pretty flexible.
Interested? My contact details are in my profile. Get in touch. I'd love to hear from you.