Curate your Data Science Journey with Web Apps
How to tell your story with simple python web apps
Purpose
The Data Science field is changing rapidly. Before, the focus was around keeping up to date with changes in machine learning libraries. Now, the focus is shifting (for the better) to domain awareness of the business problems companies face, collaborating cross-functionally with data engineers, and enabling outputs to be used across the enterprise in non-data science contexts.
All of this means that effective communication and business context go a long way in credentializing yourself in the data science field. A prime way to do this is to build a portfolio of recent work or a specific web app that tells a story for a specific problem.
For me, I went through a journey of:
- A collection of Jupyter Notebooks trying to do anything and everything data science related (Hard to share and contextualize)
- Certifications in related cloud technologies for data science (Helpful, but not an immediate pathway to a new career)
- Repositories in GitHub solving specific data science problems (Shows core skill examples but difficult to communicate)
- Web Apps demonstrating storytelling, web development, exploratory data analysis, and other skills in addition to the core Machine Learning skillset
Make it a compelling story
The more you can curate a compelling story around a data science problem to engage others, the better. Tell a story with a few select datasets that you personally find intriguing. It’s hard to build a compelling portfolio around common datasets such as the Iris dataset, but it’s also hard to a compelling portfolio with 20+ half-baked jupyter notebooks. Go deep in a few areas.
These compelling web apps don’t require years of experience in React or NodeJS. Here’s an example of a simple Flask web app I built to demonstrate so straightforward machine learning concepts.
Web App: https://python-ml-web-app-orthopedics.herokuapp.com/
GitHub: https://github.com/van-william/flask_app_ml
Examples on How to Get Started & Next Steps
- Learn a web development framework — If you’re focused on Python, pick up Flask or Django to carry over your Python experience into crafting a web app to demonstrate your data science journey
- Clean up your GitHub repositories — the more you can make your repositories cleaned up with documentation, more people can reference and fork the content
- Double down on 3–5 (max) datasets and business problems — Really go deep in a few datasets; talk to domain experts and convey your skill with communicating complex data
