Next Steps After Metis04 Apr 2016
Wow! So my 12 week at Metis is over. Here’s a recap of the last week at Metis - the whirlwind that included my final project, the career day presentation, and my next steps after Metis.
Career Day & Final Project
Career Day was last Thursday (3/31) and it’s where every Metis student presented the projects that we’ve been working toward since Day 1. It’s also a great opportunity to present in front of a large group of employers all looking to hire data scientists.
I was selected to present first, so the pressure was tremendous. My final project was about generating themed crossword puzzles using machine learning. I use machine learning to find the themes and similar clues from a large corpus of crossword puzzle clues, and then some programming elbow grease to generate the crossword puzzle.
Up until Tuesday, my project was still pretty discombobulated. I had all my separate compnents working, but haven’t been able to tie everything together yet. So the days leading up to the career day was pretty hectic. Eventually, I did get everything to work, and was even able to have a working live demo (with pre-cached puzzles) during my presentation.
Rehearsing and rehashing my presentation over and over again with other students, instructors, and our career advisor Jen was extremely helpful. I had to make a lot of changes to end up with a clear and cohesive 4 minute presentation.
I’ll be making a blog writeup of the project in the next few days, but here’s a sneak peek of the design of my full stack web-app:
Architecture stack of my crossword puzzle webapp
Friday was graduation, a very low-key ceremony where we recieved our Metis tee shirt and a certificate. Then we closed out with a nice dinner with the entire cohort. It’s bittersweet night. Sad that we won’t see each other as often, but I’m also very excited to see where everyone is heading as data scientists.
Here’s a photo of our class!
Well after a relaxing weekend, it’s back to the grind. First up, I had a technical interview scheduled with a big search company. It was the Foobar challenges that led up to this. I wasn’t very happy about my performance, so there’s a lot of room for practice and improvement.
As for next steps, I had this list of tasks and projects that I need to work on thought up in my head, but I wanted to jot them down for my personal accountability. It’s obvious that I’ll only get busier after Metis, but I’m super excited to be learning and working on them:
Scaling Up my Final Project
- Update current site of crossword puzzle generators
- Get project on Heroku
- Supporting multiple users (beyond Flask)
Updating Blog and Code
- Get all my projects and code on github and re-organize my repos
- Post blog updates on all my Metis projects
- Update blog with some additional functionality
- Write up of data science projects I’ve worked on outside of Metis
- Reach out to employer connections through Metis
- Apply to 2+ jobs a day
- Attend meetups and other networking opportunities
Brushing Up and Reviewing Data Science
- Finish Andrew Ng’s Machine Learning course on Coursera
- Then go thrugh NLTK Book on NLP
- And then go through Elements of Statistical Learning
- Review and implement data sturctures and algorithms
- Maybe even applying them to data science related projects
- 3 programming challenges everyday
- Project Euler
- D&D resume on website
- Slack bot to cast spells
- Mapping rats in NYC (splinter and TMNT)
- Finding haikus in Python (Pyku)
Continue Kaggle Competition
For project 3 at Metis, I worked on a Kaggle competition to classify insurance claims. This competition is still open for about 2 week, so I will be working to improve my model and my ranking.
Consulting on the Side
I may also have a chance to work with an old friend to apply data science for his hospitality business. I’ll be discussing the project with the company in the next few days, but this would be an excellent opportunity to apply my skills to help a real business.