Open to opportunities
Hey, I'm Ananthamoorthi
I build full-stack AI products that solve real problems
LLM systems that hold up in production: grounded retrieval, validated output, and multi-agent workflows. Founding engineer at SutraAI Solutions, where I took mySutra.ai from ideation to its Google Play Store launch.
Python•
JavaScript•
TypeScript•
React•
Next.js•
Node.js•
Flutter•
FastAPI•
PostgreSQL•
Supabase•
Firebase•
Express.js•
Docker•
Tailwind•
Git•What I bring to the table
A blend of technical depth and product thinking to build solutions that matter.
AI Engineering
LLM orchestration, RAG, multi-agent LangGraph workflows, evaluation, and guardrails. Shipping AI features that stay reliable in front of live users.
Backend Development
Production backends with Python and FastAPI. REST APIs, data models, and async workers serving live applications.
Database Architecture
Designing data systems with PostgreSQL, pgvector, ChromaDB, Supabase, and Firebase for retrieval-heavy applications.
Product Engineering
End-to-end product thinking, from ideation to deployment, focused on what users actually need rather than what demos well.
Web Development
Crafting modern web apps with React and Next.js that are fast, responsive, and built to scale.
Mobile & On-Device AI
Cross-platform apps with Flutter and React Native, including quantized models running fully offline on the phone via llama.cpp.
Take a look at my
project portfolio
mySutra.ai, AI Health Intelligence Platform
Photograph a meal, and the app works out what you ate and turns it into insight you can act on, with an AI coach that reads the trend rather than the number. I owned the AI side from first prototype to launch, built the backend it runs on, and shipped the app to the Play Store.
Visit website
Project Sentinel, Autonomous Cold-Chain Monitoring
Refrigerated medicine spoils quietly, and by the time anyone notices the temperature slipped, the shipment is gone. Sentinel watches shipments in real time and reroutes them to cold storage on its own, through a five-stage LangGraph pipeline. The final commit is deterministic scoring rather than the model, so every reroute can be reproduced and audited.
View on GitHub
PlainLabs, Offline Lab-Report Explainer
Upload a blood report and get every value explained in plain language, flagged normal, borderline, abnormal or urgent, with questions worth asking a doctor. Runs fully offline on a 4B on-device model. The model only writes the words; every severity flag is a deterministic comparison against curated reference ranges.
View on GitHub
Doc Companion, Agentic RAG for PDFs
A LangGraph agent that decides how to answer each question instead of retrieving blindly, and cites the page it came from. It grades its own chunks, rewrites the query when retrieval misses, and answers corpus-level questions from per-document cards rather than top-k. Runs fully offline with an eval harness scoring faithfulness.
View on GitHub
LocalGPT, On-Device AI Chat
Quantized LLMs running entirely on the phone through llama.cpp. No network, no data leaving the device, with GPU offload where available to keep latency usable on mid-range hardware.
View on GitHub
Take a look at my work experience
From AI-driven product development to full-stack applications, I've built scalable solutions with a focus on performance, usability, and real-world impact.
AI Engineer, Founding Team
@ SutraAI Solutions LLP
Took mySutra.ai from ideation to its production launch on the Google Play Store. Built the AI backend end to end with multi-step LangGraph workflows, a RAG pipeline grounded in verified nutrition data, and guardrails for medical-advice boundaries validated by retrieval-quality evaluation. Also built the core backend and shipped the Flutter app.
Project Intern
@ IoTracX, Manipal
Worked on on-device LLM inference for offline use, tuning latency and memory footprint for constrained mobile hardware. Ran Visual Language Model and fine-tuning experiments, and refined datasets to improve training data quality.
Intern, IT Department
@ Wipro Consumer Care & Lighting
Contributed to internal IT prototypes for a Market Survey application and a Vendor Portal, working on Node.js backend services and React frontend components. First hands-on exposure to enterprise practices: requirement discussions, version control workflows, and iterative delivery.
Who's behind all this great work?
I'm K. Ananthamoorthi Holla, an AI Engineer with 1+ year of production experience building LLM systems end to end. As founding engineer at SutraAI, I took mySutra.ai from ideation to its launch on the Google Play Store, designing the AI backend, building the core backend, and shipping the Flutter app. A model is good at language. It shouldn't be the thing that decides whether a number is dangerous, and most of what I build is some version of that split.
B.E. in AI & Machine Learning
Visvesvaraya Technological University (VTU), Belagavi, with CGPA 8.4 (2021-2025).
Reliable AI, not just clever AI
RAG and multi-agent LangGraph workflows, LLM evaluation, guardrails and hallucination control, fine tuning with Unsloth, and running models fully offline on device.