Watchful is an ML seed stage startup that speeds up Full Text Classification data labeling by an order of magnitude. Watchful achieves this by encoding a user’s concepts of membership (ex: Sports articles often contain the word ball) into heuristics used to create probabilistic hand labels across a user’s dataset.
▪ One of two front-end engineers to build a V2 User Interface to pave the way for new data modalities, integrations to third
parties, and labeling tasks like Named Entity Recognition. Owned the epic for the final month: coordinated work for 5
engineers, decided on readiness to ship, reported directly to the CEO.
▪ Created the frontend auth-flow necessary introduce authenticated users across the Watchful ecosystem:
◦ Collaborated with backend counterpart to design and execute a RESTful authenticated API, and built the front end for Watchful’s user and admin console, Watchful Hub
◦ Created all user related interfaces in the watchful app, and gated authenticated functionality.
◦ Created and documented front end auth flows for both the Watchful application and Watchful Hub
▪ Increased engineering velocity by building a dev flow to mock API calls with high generality, allowing engineers to
document desired API behavior in code.