Project Overview
- Everything built so far in the data and analytics pipeline led to the envisioning and development of this engine
- As the product owner, I performed feasibility analysis, prepared PRDs and even dashboard mockups using Figma
- As the developer, I was responsible for the data, ML and backend components, involving:
- Building self service data APIs and data for visualization APIs
- Building FastAPI endpoints for models with poetry dependency management, dockerizing them and deploying them at scale on ECS
- Cognito, Aurora and Okta SSO integration for single sign on, alongside user token and session management for security
- Entire platform and environment planning and setup, along with developer account(s) and IAM service roles management
- As the team lead, I was also responsible for overlooking the design and development directions and progress of a third party front end team
Skills
MLOps
While MLOps was not new to me, this was the first time where I had to plan and implement the entire operational aspects, alongside development
Total Product Development
The amount of exposure gained in building this product gave me an idea of all the modules needed to build a product from scratch and make it ready to host thousands of users. And the opportunity to develop almost all those modules except frontend, strenghtened that knowledge and ability
Product Management
With the involvement of a third party development team and with natural occurences of dependencies and scope revision, I also had the opportunity to perform product management and develop those skills