I started in Mechanical Engineering, but systems thinking pulled me toward software — and four years later, I'm still chasing the same question that got me hooked: how do you build something that actually holds up in production, end to end? At Incresco, that question took me across very different terrain. I architected a unified multi-app CRM platform — stitching together CRM, Support, Admin, Inventory, and CMS under one roof with seamless shared auth — and then pushed it further with a LangGraph-based multi-agent AI system that let teams query their own tenant data in plain language. On the client side, I took a booking platform and rebuilt it into a hotel discovery product, migrating over a million customer records and 100GB of data to a new CRM and search infrastructure. I also built an internal HRMS that let employees handle leave, payroll, and reimbursements straight from WhatsApp, because the best tools are the ones people don't have to think about using. Now I'm at Ztal.ai, building AI infrastructure for staffing firms — semantic search and AI-driven screening that helps recruiters find the right person faster. It's a different domain, but the same instinct: take something messy and manual, and make it fast, reliable, and almost invisible. I lean heavily on AI-Assisted Coding with Codex and Claude to move quicker without cutting corners — pairing that with React, Python, and Pydantic on the daily. I care about owning things fully — schema to deployment, not just the parts that look good in a demo. Philosophy: Clarity beats cleverness. I build systems that scale, document so knowledge compounds, and debug what I don't understand before shipping it. Right now, that means designing AI systems responsibly — not just wiring up models, but understanding what they get wrong.