Designing robust AI systems. Balancing neural execution with deterministic control.

AI Full Stack Engineer specializing in agentic workflows, LLM-powered automation, and backend systems built for production. I design architecturally sound solutions, map out solution boundaries, and build low-latency API pipelines across Python, Node.js, and TypeScript.

Analyze
(break down)
Understand
(architecture)
Research
(tools & data)
Build MVP
(test & launch)
Optimize
(iterate & scale)

About Me

I design and build software at the intersection of AI models and backend infrastructure. My work centers on creating multi-agent setups, optimizing vector search indexes, and deploying deterministic state machines to control unstructured AI outputs.

Currently at LogicKrew, I build observability tools, model evaluation pipelines, and persistent agent memories. Before that, I launched Lumina, a native macOS and Windows application supporting clinical research automation.

Availability: Based in India | Open to remote roles and relocation for long-term opportunities.

Featured Projects

Production systems built end to end.

LocationHQ (VDLHQ)

Active
Python · FastAPI · pgvector

A production operations platform that automates lead workflows and uses vector similarity to match film sets with property profiles.

⚡ 70% load reduction | P99 < 85ms | 98.5% precision
View Case Study

AI Lead Intelligence & Routing System

Completed
n8n · Local LLMs · Gmail API

An event-driven lead intelligence processor that scrapes corporate details, scores urgency, and generates draft replies.

⚡ 40% speedup | 95% LLM accuracy
View Case Study