I build AI agents, RAG systems, and intelligent automations that connect LLMs with APIs, databases, and business workflows — so teams stop copy-pasting and start shipping.
I build intelligent systems that go beyond chat — AI that retrieves information, makes decisions, calls APIs, and completes real business tasks with minimal supervision.
My focus is Agentic AI, Retrieval-Augmented Generation, workflow automation, and applied machine learning. Python and SQL are my daily tools.
I've built my foundation through programs from IBM, Stanford, DeepLearning.AI, the University of Michigan, and UC Davis — which gave me hands-on experience across machine learning, databases, AI agents, and automation.
I'm at the beginning of my professional journey and I'm honest about that. What I bring instead is speed, clean code, and the discipline to ship things that actually work. Consistency and execution matter more than years on a résumé.
No buzzword salad. These are the tools I use to ship agents, retrieval pipelines, and automations end-to-end.
Focused builds that demonstrate how I approach real business problems with AI — retrieval, reasoning, and action.
RAG-powered assistant that actually reads the docs.
Support teams waste hours answering questions already documented in PDFs and help centers.
A conversational agent that ingests a PDF knowledge base, retrieves the right passages semantically, and answers in context — with memory across turns.
Turn 60 minutes of talk into 60 seconds of clarity.
Meeting transcripts are long, noisy, and rarely acted upon.
Upload a transcript; the model extracts action items, decisions, and a crisp summary — served through a clean REST API for downstream tools.
Ask in English. Get validated SQL and results.
Non-technical teammates can't query the database without engineering support.
Natural language → schema-aware SQL → safe execution against a sandboxed database, with results returned in plain language.
Classify, route, and draft — before you open the inbox.
Founders drown in email triage that eats the first two hours of every day.
Incoming emails are classified by intent, routed to the correct workflow, and given a draft reply that a human just needs to approve.
See the world's stories, grouped and summarized.
Following global events across dozens of sources is impossible manually.
Articles are collected, embedded, clustered by story, and summarized — with semantic search over the archive.
Every system I build follows the same loop: understand, retrieve, decide, act. This is what that looks like in production.
Certifications from teams that literally wrote the ML playbook — grounded, not decorative.
Engagements are scoped tight, delivered in weeks — not quarters.
New APIs, new frameworks, new domains — I close the gap quickly and honestly.
Regular async updates. No black boxes. You always know where the project stands.
I optimize for outcomes that ship, not demos that impress.
Scoped work, agreed timelines, and follow-through until it's live.
Readable, typed, and structured so future-you (or your team) can extend it.
The frontier moves weekly. So do I.
"Client testimonials coming soon as I complete freelance projects."
Want to be the first case study? Let's talk.
Tell me about the workflow you want to automate or the assistant you want to launch. I reply within 24 hours.