Portfolio · 2026
Data Scientist & AI Engineer.San Francisco Bay Area · currently at INVIDI.
About
I've always enjoyed building things, from scribbling logic in notebooks to shipping code that powers real products. Somewhere along the way, I got curious about the patterns behind the data. That curiosity led me into machine learning, and eventually into designing systems that don't just run, but learn.
These days, I work at the intersection of software engineering and intelligence, building backend services, search systems, and ML pipelines. Lately I've been deep into semantic search and autonomous agents, and anything that turns raw data into real insight.
I'm based in the San Francisco Bay Area. Outside of work, you'll usually find me hacking on side ideas or chasing good light with my camera. If you're curious or want to collaborate, reach out.
Skills
Designing agents that use tools, hold memory, and act in loops without falling over past the demo.
Embedding pipelines, hybrid search, and the eval loops that keep retrieval honest.
Taking models from notebook to deployed APIs with infra that doesn't rot under traffic.
Experience
A chronicle of the systems, teams, and problems I've shipped against.
Data Science & ML
Leading data science initiatives focused on forecasting, optimization, and MLOps. Built MLOps pipeline using AWS SageMaker and API Gateway to deploy ML models as scalable, serverless endpoints. Currently building AI-powered semantic search systems for video content discovery and matching.
Software Developer Intern
Developed advanced inventory scheduling system for efficient ad campaign delivery. Achieved 98% utilization rate of ad inventory.
Senior Software Engineer
Developed enterprise applications and ML-powered features for Visa's internal platforms. Improved retention and user experience by 40%.
Backend Engineer
Developed backend solutions for major international clients including Kia Motors and InterContinental Hotels Group.
Associate Technology
Developed web solutions and ML models for clients and internal projects. Created analytics for marketing campaigns and customer profiles.
Education
Masters in Computer Science
Advanced studies focused on machine learning, AI, and data science. Research Assistant on a longitudinal study of local news in New Jersey.
B.Tech, Electronics & Communication
Built a strong foundation in engineering principles, programming, and signal systems. Developed alcohol-detection-based accident prevention technology as final-year project.
Projects
Hackathons, side ideas, and a few things I built to learn how something actually works.
Autonomous AI agent platform for activity tracking — uses computer vision, LLMs, persistent memory, and adaptive real-time decision-making.
Privacy-first AI meeting assistant — local Whisper/Parakeet transcription on-device with no bots joining your calls, dual audio capture, GPU acceleration on Metal/CUDA/Vulkan, and your choice of AI provider for summaries. Tauri desktop app for macOS and Windows. Currently shipping as a SaaS.
End-to-end voice-AI recruiting platform. Multi-state voice interview agent over web and phone, with silent mid-call extraction that compresses conversation into structured signal before the closing summary. Hardened with a layered prompt-injection defense for untrusted candidate input. Backed by hybrid semantic search (vector + BM25), LLM rubric scoring with must-have/preference classification and tech-adjacency rules, LinkedIn enrichment, and a full pipeline CRM.
Chrome extension using FalconLLM and LangChain for 60-word web content summarization, with an interactive chat layer.
Containerized, microservice-based recommendation system on FastAPI + Redis + Docker + Kubernetes with autoscaling and caching.
Retrieval-augmented document QA on LangChain + HuggingFace + FAISS with efficient chunking and reranking.
StyleGAN from scratch on FFHQ with Few-Shot GDA via Domain Re-modulation (DoRM) for cross-domain adaptation.
Spiking neural network for sign-language recognition on the ASL Dynamic Vision Sensor dataset — 96.7% test accuracy.
Global socioeconomic patterns in suicide trends — GDP correlation analysis with age/gender breakdowns in R.
Contact
Book a coffee chat or send a note — whichever's easier.