Coding Projects

A list of all my coding projects ordered by project size.

Currently looking for a FinTech or Data Science job. Contact me via LinkedIn.

I have a business-tech background. You can find more info on my CV.

Noisy Stocks

Big data application which automatically generates spurious stock correlations between stock price and rainfall based on historical movers & shakers.
Published online on noisystocks.com.
Orchestrator code & additional info on Github.
Features:
- Blazing fast. Can process 60+ stocks to a fully published state in less than 5 minutes.
- Automatically finds correlations based on time series
- Optimized Big Data database
  - more than 870 million weather & 14 million stock datapoints
  - 90+% compression rate
- Decouples back-end & front-end securely using Jamstack architecture.
- Strongly typed Python using Pydantic & Pandera to increase reliability
- Fully idempotent. No matter how many times you run the program, the result will be as if it only ran once. A very useful property in case your computer loses power suddenly.
- DevOps best practices:
  - Continuous Integration
  - Continuous Deployment
  - Reproducable development environment