BarLens
Production applied AI work focused on reliable product workflows, human review, and measurable user outcomes.
Portfolio
Detailed writeups with links, stack, and impact.
Production applied AI work focused on reliable product workflows, human review, and measurable user outcomes.
ARI module that turns a candidate's take-home project into a structured, role-specific interview brief.
Agent-oriented learning platform that turns uploaded materials into personalized study outputs with teacher oversight.
Capability broker that routes agent tasks to the right tool, redacts sensitive context, and gates paid execution.
End-to-end TTC reliability platform covering 1,004,682 delay events across bus, subway, GTFS, and live-alert data.
Research-driven project exploring LLM + graph representations for constrained optimization (JSSP and related tasks).
Anomaly detection on urban traffic streams comparing probabilistic GLM and sequence-model (GRU) approaches.
ROS2 perception-to-manipulation pipeline using YOLOv8, OpenCV transforms, and ArUco calibration for precise robotic sorting.
Planning assistant to structure exam preparation timelines and task prioritization.
Fine-tuned multilingual transformer summarization and translation pipeline for legal text, packaged as a Flask service.
Retrieval-grounded document QA system using LangChain and FAISS with agentic query workflows.
Semi-automated grading system with BERT-based NLP and OCR to evaluate typed and handwritten responses.
Comparative ML and deep learning study for Uber NYC driver pay prediction with analysis dashboards.
Regression modeling and feature analysis pipeline for India weather forecasting competition data.
Natural-language interface for uploaded SQL databases, translating prompts into executable SQL queries.
Android Kotlin app using Reddit APIs and Glide for lightweight meme streaming UX.