Designing an AI-powered journaling interface using React and TypeScript, allowing users to submit entries and receive LLM-generated summaries and affirmations within 3 seconds, resolving UI defects, reducing reflection time, and boosting journaling frequency by 40%. Implementing PostgreSQL-backed data storage to persist journal entries, LLM outputs, and sentiment scores, enabling consistent emotional tracking and seamless history retrieval across 100+ test users.

Conducted Python lab sessions for 90 students, teaching advanced concepts such as decorators, multithreading, collections, and REST API integration to support real-world application development. Graded assignments and participated in project presentations, providing constructive feedback that enhanced students' understanding of project development and promoted the effective application of advanced Python concepts.

Designed and implemented an interactive frontend in React and TypeScript for case uploads, prompt selection, model execution, and leaderboard visualization, enabling structured evaluations and cutting manual review time by 35%. Developed RESTful APIs in Python (Django) with a PostgreSQL schema for datasets, prompts, runs, and results, ensuring consistent scoring and reproducible comparisons across 1,200+ oncology clinical notes. Automated oncology field evaluations by integrating LLM adapters (ChatGPT, Claude, Gemini, LLaMA) with strict JSON schema validation, delivering audit-ready comparisons and accelerating model selection cycles by 5×. Led cross-functional Agile SDLC activities, driving weekly planning, backlog grooming, standups, and review sessions with researchers and engineers to align priorities, remove blockers, and maintain delivery momentum.

Built scalable back-end APIs using Python (Django) for an Auto Label Comparison Tool, implementing authentication, data validation, and ingredient analysis features, streamlining identification of inconsistencies and improving label accuracy by 70%. Diagnosed and resolved data handling bugs in the label comparison pipeline by analyzing inconsistent API responses and applying logging and debugging tools, enabling HIPAA-compliant processing of over 20M label records and reducing compliance risk. Optimized SQL queries and stored procedures to process, validate, and reconcile large-scale label data, reducing query processing time from 120 seconds to 70 seconds, accelerating label comparison workflows and supporting reliable downstream analyses.

Implemented Python (Django) APIs to aggregate and serve satellite anomaly data from sources like Next Launch, applying Redis caching and optimizing data retrieval logic, achieving faster, more reliable access to critical information while reducing external API calls by 60%. Enhanced modular UI components using Angular and TypeScript, transforming complex satellite anomaly datasets into intuitive dashboards and visualizations, boosting data analysis efficiency 4x and increasing user engagement. Led Agile development within the Software Development Life Cycle, mentored junior developers, improved delivery clarity through structured documentation, oversaw deployments to production environments, and reduced production bugs by 120+ incidents per quarter.