Open Deep Research
A modular, adaptable deep research framework built for enterprise. Autonomous agents plan, search, scrape, rank, and synthesize cited intelligence — designed so teams can plug in new domains without rewriting the core.

Engineer first — also an entrepreneur and consultant. I build thoughtful systems, and I like working with people who care about getting it right.
Based in New York — available in person across the NYC metro area, and remote anywhere.
I founded Luminary AI Solutions to help teams ship agentic research and production software that feels dependable — not flashy for its own sake. Whether it's an open-source research framework or a school management platform in the wild, the goal is the same: clear thinking, careful craft, and work you can trust.
Full-stack across Python and TypeScript, with a special focus on AI frameworks that hold up in production.
From open-source research systems to enterprise software in pilot — a mix of product, client work, and experiments.
A modular, adaptable deep research framework built for enterprise. Autonomous agents plan, search, scrape, rank, and synthesize cited intelligence — designed so teams can plug in new domains without rewriting the core.
Production-grade school management software built to contract. Payments, registrations, attendance, academics, calendar, and notifications — a sophisticated system design now piloting before wider licensing.
Demo or code walkthrough available by request.
“Before we began working with Luminary on the Maktab software, our school management and registration process were in disarray. We were losing thousands of dollars per year chasing down registration fees. Working with Luminary has been a great experience from start to finish. They are strong communicators, focusing hard on our needs, and delivering top quality work with an impressive turnaround time.”
Intelligent retrieval across large document sets with specialized search methodologies — a practical foundation for grounded, citation-aware AI products.
Agentic market analysis that conglomerates fundamentals, earnings calls, news, social sentiment, and financials — then analyzes and synthesizes with RAG and a chatbot over the data.
AI-powered patient nursing request system — an MVP built for investor pitching, with a patient-first interface that goes well beyond traditional call-bell workflows.
“Abdullah from Luminary AI blew me away with his helpfulness and depth of expertise in AI systems and applications. I've worked with him on two projects now, Moodi, an AI powered patient care app, and my latest MVP, Forward Care. Throughout the process he was timely, knowledgeable, and helped me navigate the complex AI ecosystem excellently.”
A B2C app for seamless event photo sharing — serverless architecture built for speed and simplicity when moments matter.
Machine learning dashboard for non-generative sentiment analysis on earnings calls, paired with traditional data analysis to surface signal for investors.
Technical depth matters — so does being someone teams enjoy building with.
Written and oral — equally at home explaining architecture to engineers or outcomes to stakeholders.
Keeps complex builds on track: scopes, timelines, and the details that make delivery feel calm.
Comfortable owning direction, coordinating people, and holding a high bar without the noise.
Experience selling and delivering to enterprise clients — listening first, then building trust.
A nerd for current technology, and capable of reading research papers when the problem demands depth.
Teaching on weekends — patience, clarity, and showing up for community.
Leading and organizing locally — the same care that shows up in client work.
Strategy, patience, and the long game — on the field and in the stack.
Whether you need a research system, a production product, or a thoughtful second opinion — I'd love to hear from you.
Get in touch →