I build Agentic AI systems
that ship real outcomes.
About
Agentic AI Architect | AI Engineering Leader | Distributed Systems
I’m an engineering leader and AI architect with 13+ years of experience building large-scale distributed systems and production-grade AI-driven automation. I specialize in Agentic AI platforms, multi-agent orchestration, and Retrieval-Augmented Generation (RAG) systems—turning enterprise knowledge and workflows into reliable, measurable automation.
In my current role, I design and deploy AI orchestration layers (LangGraph + Python + FastAPI), build RAG platforms with vector search and semantic retrieval, and deliver autonomous agents that accelerate engineering and operational workflows—alongside scaling high-throughput APIs serving millions of requests per day.
Current Focus: Agentic AI platforms // RAG & enterprise knowledge systems // AI-powered developer productivity
Experience
Agentic AI platforms, RAG systems, and high-throughput APIs
Built production AI orchestration and large-scale platforms
- Architected and implemented Agentic AI systems using Python, FastAPI, and LangGraph to automate engineering and operational workflows.
- Designed and deployed 20+ autonomous agents for productivity improvement, internal tooling, and workflow automation.
- Built RAG platforms integrating vector databases, semantic retrieval, embeddings, and enterprise knowledge bases for contextual automation.
- Scaled APIs handling 3M+ credit inquiries/day with sub-second response times and thousands of real-time requests per minute.
- Led and mentored hybrid teams (15+), spanning backend, platform, and AI initiatives across Ruby on Rails, Elixir, and Python on AWS.
Ops portal that enables copart operations team to build, run, and measure vehicle life cycle
Develop Product
- Built scalable APIs and backend systems supporting global automotive auction and logistics platforms.
- Led migration of legacy systems into modern distributed services; improved reliability and performance for high-traffic workloads.
- Developed workflow automation systems reducing operational overhead and improving system efficiency.
- Tools: Ruby, Rails, MySQL // React, JS, HTML, JSX, Haml
Accomplishments
- Dramatically increased speed of DMV project development using service based design
- Built tool to scrate data from DMV, MVR sites to simplify yard inventory
- Implemented inline @tagging feature to increase business reporting
SaaS platform helping hundreds of Iindian businesses manage their employees
- Back-end Ruby on Rails server, with HTTP & WebSockets endpoints, and MongoDB, Redis & ElasticSearch databases.
- Implemeted Quikchex V2 product from scrath and helped in team growth
- Worked with business & operations team for requirement gathering and analysis
Accomplishments
- Improved system performance using caching & memorization
- Moved Monololithic application into web services
- Wrote an app to auto sync employee attendance records from various finger print devices. Quikchex was the only product in India with this feature.
- Built an open source Email generator app to send marketing emails, which can send millions of emails per day.
Sass product for Educational Institutions
- Built online exmaination app uisng Ruby, rails, HTML, CSS, jQuery.
- Production support for Inbot app.
Skills
AI / GenAI
AI Architecture
Back-End
Javascript
HTML
CSS
Other
Awards
Copart India Technology Center | April 2019
- Awarded by Vice Precident of Engineering in the Copart to one engineer as recognition for "design and implementation of DMV & MVR projects and created framework/foundation to add new states with minimal code & for mentoring juniors and communicating with everyone including business."
Copart India Technology Center | May 2019
- Quarterly award given to a team member who exemplifies Copart's purpose, vision, and values, goes above and beyond in their role to make a particularly positive impact on the company.
Copart India Technology Center | November 2018
- Nominated by management and coworkers for exceptional work ethic and dedication for G1 2.5 project.