Currently building at goML

Building AI systems
that move from experimentation
to production.

I design inference systems, agent tooling, evaluation pipelines, and developer experiences that help teams ship reliable AI products.

See selected work Get in touch

Coimbatore, IN — 14:22 IST

Role

MTS, Applied AI

goML

Experience

2 years

Certifications

2× AWS Certified

India · Remote-friendly

Built & shipped

Technical articles

Years building

2×

AWS Certified

Contributions

About

A story,
not a résumé.

I came to AI through the part of engineering that feels closest to craft — making opaque systems behave predictably, and giving people a clear surface to use them. The model is rarely the interesting problem. The interesting problem is the loop around it.

Most of my work lives between research code and production: inference pipelines that have to stay up, agents that have to fail safely, retrieval systems that have to be auditable. I care about the seams — evaluation, observability, prompts as configuration, the deploy story.

“I optimize for systems that people actually use — not demos.”

I build products because the constraints are honest. A feature ships or it doesn't. A latency budget is met or it isn't. That feedback loop is what keeps the work serious.

Work

Where I've been shipping.

Jun 2025 – Present

Coimbatore, India

Active

Designed production-grade EHR AI pipelines integrating relational, document, and vector databases.
Built multi-query retrieval architectures for structured healthcare data extraction.
Engineered AI travel booking agents with Explore/Discover routing strategies.
Integrated Amadeus and Sabre GDS systems with intelligent routing and silent failover.
Built a seven-agent orchestration framework with pgvector, semantic retrieval, sentiment analysis, dual-memory architecture, and recommendation pipelines.
Worked across backend architecture, evaluation systems, deployment infrastructure, prompt engineering, and production AI delivery.

Jun 2025 – Jun 2026

Coimbatore, India

Contributed to internal AI platforms by developing foundational agent workflows, evaluation pipelines, inference services, deployment automation, and developer tooling that supported production LLM applications.

Nov 2023 – Jul 2024

Coimbatore, India

Developed LSTM and SVM spam detection models achieving 97–98% classification accuracy on 50K+ emails.
Built retail analytics pipelines that identified sales trends and contributed to a 25% improvement in inventory turnover.
Worked on end-to-end ML workflows including data preprocessing, feature engineering, model evaluation, and deployment.

Sep 2023 – Nov 2023

Coimbatore, India

Built a graph-based recommendation engine using NetworkX.
Improved recommendation accuracy by 15% across more than 10,000 user interactions.
Increased user engagement by approximately 20% through personalized recommendation strategies.

Selected projects