I build AI systems that work in production, and the teams and products around them.
Not a list of buzzwords. A working system across every layer of production AI.
End-to-end AI pipeline for automated invoice data extraction. Handles unstructured PDFs, handwritten documents, and multi-format invoices with high accuracy using LLMs + custom extraction logic.
Intelligent document processing system for mortgage workflows. Validates, classifies, and extracts structured data from complex legal documents, cutting processing time and compliance risk.
Full LMS built for a tech academy in Kerala. Includes live portfolio tracking, ATS-optimized resume generation, progress dashboards, and AI-assisted curriculum delivery for 100+ active learners.
Multi-agent system that automates complex business workflows: from research and data gathering to decision-making and reporting, using LLM agents with tool use and memory.
Products I built, launched, and own. Hobby projects.
A lightweight, fast CRM built for sales teams who are done with spreadsheet chaos. Zero-code customization, role-based visibility, and real-time performance insights. The CRM your team will actually use.
An AI sales call assistant embedded inside LynkLead. Swarra transcribes recordings in 99+ languages, extracts key moments, surfaces objections and commitments, and posts structured summaries directly to lead timelines.
The difference between a prototype and a product is how you think, not just what you code.
Dig into the real problem. Most requests are symptoms. I find the root cause before writing a line.
Design the system before touching code. Data flow, failure modes, scalability: all mapped upfront.
Ship working software fast. Iterate in small cycles. No gold-plating. Value delivered early.
Production-grade from day one. CI/CD, monitoring, error handling: not an afterthought.
Performance tuning, team scaling, architecture evolution. Built to grow without breaking.
I'm an AI engineer who builds things that matter in production, not demos. My core strength is bridging the gap between cutting-edge AI and real business outcomes: fast, scalable, and deployed.
I'm currently a GenAI Engineer at Evolver (Palo Alto, remote), and have previously built production AI systems at Perfecter.ai and led data science work at Dexpro Innovations. I also hold stakes in BlackBox Learning and PRception.
Beyond writing code, I architect teams, lead cross-functional builds, and drive product strategy from ideation to deployment. If you're building something ambitious with AI, I want to hear about it.
Engineering is half the picture. Here's the rest.
Whether it's a production AI system, a product from scratch, or a team that can execute, I'm interested. Let's talk.