At Ethara AI, I evaluated and debugged LLM-generated Python code and reasoning outputs, checking syntax, runtime logic, and instruction following. I also analyzed edge cases and provided structured feedback for model-alignment pipelines.
I curated dataset pairs for reasoning, multimodal, and text-to-code workflows, and verified visual model outputs against their input constraints.
On SnapMenu, I developed a Python backend with Flask and SQLAlchemy to manage order and transaction data. I implemented workflows that transformed daily transaction data into PDF business reports and helped automate reporting.
For my Facial Recognition & Attendance System project, I developed a Flask REST API to exchange frame-inference data through JSON. I worked on integrating computer-vision inference with application-level data exchange and investigated pipeline latency and error handling.
