AI Engineer
Captus.aiBuilt and deployed production-ready AI agents and automated workflows, and designed evaluation pipelines that boosted agent accuracy from 20% to 80%.
Built and deployed production-ready AI agents and automated workflows, and designed evaluation pipelines that boosted agent accuracy from 20% to 80%.
Mentoring students in Python, automation, social media ethics, and data mining.
Experiments and analysis of social media algorithms topic.
Mentored first-year Informatics students as a FIG Leader/Mentor, guiding campus life, academic, and technical development.
Organized 20+ events, secured $5K in sponsorships, and mentored 25+ members.
Built a personal healthcare assistant agent integrated with iMessage using the Photon iMessage SDK. Users activate the agent by texting, which gathers symptoms and context through follow-up questions, then provides actionable next steps like nearby in-network clinics, questions to ask your doctor, and pre-visit care recommendations.
Built a competitive game-playing agent using Minimax with Alpha-Beta pruning and heuristic optimization. Integrated LLM-based response generation for adaptive, context-aware agent behavior.
Designed a gamified coding platform for ADHD learners, combining real-time feedback, accessibility-first UI, and progress tracking. Implemented Firebase-backed personalization for adaptive learning paths.
Built a Java-based maze visualizer with shortest-path solving using Dijkstra's algorithm, achieving 40% faster computation. Implemented Kruskal-based maze generator to produce fully solvable mazes at scale.
Built a recipe search and recommendation system with a clean web interface. Developed a Python web crawler to collect 5,000+ recipes. Implemented BM25, TF-IDF, and LM-based retrieval models.
Here's some music I've been liking recently
Drag a record onto the player