Nakul Gupta / Founding AI Engineer
I design reliable systems for messy data, high-stakes workflows, and the people who depend on both.
Two bodies of work that show how I think: trace the data, control the expensive path, and design for operators.
A traceable pipeline that turns enterprise inboxes and documents into structured case intelligence.
Read case studyNon-blocking workspace operations across a Rust terminal UI and React/Tauri desktop app.
Read case study
Founding AI Engineer
Production legal AI, data architecture, and customer delivery across intake, document intelligence, retrieval, and multi-tenant systems.
Machine Learning Research Assistant
Built Python data-processing and ML analysis pipelines for logistics operations, transforming forklift maintenance data for mediation analysis and business-facing insights.
Software Engineer
Built Java and Spring Boot ingestion services for industrial IoT workflows, processed 1M+ NEOM data points, and designed a retention plugin that reduced storage overhead by 60% while mentoring junior engineers.
I am an AI engineer who moves between architecture, implementation, and customer discovery. At Theo AI, I have worked on intake pipelines, document intelligence, graph-backed retrieval, multi-tenant data systems, and the operational controls that make AI dependable.
I care about the unglamorous parts: provenance, retries, cost accounting, migrations, access boundaries, and interfaces that make complex systems legible.