About
Building backend & GenAI systems that scale
Senior Backend & GenAI Engineer with 7+ years building scalable, cloud-native systems in Node.js and Python. Specialized in LLM integration, GenAI workflows, and healthcare interoperability (FHIR, CommonWell, EHX). Experienced in AWS serverless architectures, multi-cloud deployments, and leading backend teams through complex migrations — with a growing focus on applied AI and agentic systems.
Tech stack
Skills & technologies
Languages
Node.jsPythonTypeScriptJavaScriptBash
Backend
REST APIsMicroservicesGraphQLExpress.jsFastAPIEvent-driven
Cloud / AWS
LambdaSQS / SNSECSS3EC2CognitoStep FunctionsAzure
Databases
PostgreSQLMySQLMongoDBRedisElasticsearch
Healthcare
FHIRCommonWellEHXHIPAA
GenAI / LLM
LLM IntegrationLangChainPrompt EngineeringAWS BedrockAgentic WorkflowsIBM WatsonMicrosoft LLMRAG
DevOps
DockerCI/CDTerraformGitLab
Career
Work experience
July 2024 — Present
Senior Software Engineer
Trove Health — Hyderabad, India
July 2023 — July 2024
Senior Software Engineer
Innovatics LLP — Gujarat, India
Jan 2022 — July 2023
Senior Software Engineer
Bit01 Technopole Limited — Gujarat, India
June 2017 — Jan 2022
Senior Software Engineer
Telenetix Private Limited — Udupi, India
Work & side projects
All projects
Patient Access & Healthcare Interoperability Platform
2024 — Present
Node.jsAWS SQS/SNSLambdaFHIRCommonWellEHX
- Integrated backend services with CommonWell and EHX APIs for patient access, identity matching, and clinical data retrieval
- Migrated systems to CommonWell v2 APIs, improving patient identity matching accuracy by 25%
- Coordinated end-to-end migration of 3 AWS-hosted databases to Azure MySQL — completed in 1 week with zero downtime
- Authored 8–10 reusable NPM modules to enhance code modularity and cross-project consistency
- Integrated FHIR resources (Patient, Encounter, DocumentReference, Practitioner) for standardized healthcare workflows
- Used AWS SQS and SNS for async messaging, notifications, and workflow orchestration
Inventory System for Pharmacy / Labs
2023 — 2024
PythonElasticsearchDockerAWS S3 / ECS
- Built Medicine Master search with Elasticsearch — achieved 90% faster search vs previous implementation
- Containerized with Docker and established CI/CD pipeline, making deployments 100% faster
- Redesigned Invoice and Billing as a standalone microservice, reducing 70% of rework
NFT Mantra — Blockchain Backend
2022
Node.jsEthereum / PolygonTatum KMSIPFS
- Architected backend services enabling users to upload, mint, and list NFTs on Blockchain Mainnet
- Achieved 50% increase in user engagement and 40% growth in daily transactions
- Built Tatum Key Management System for secure wallet private key management
Antonic AI — Conversational Platform
2022
Microsoft LLMLangChainIBM Watson
- Leveraged Microsoft LLM and LangChain for advanced natural language understanding
- Integrated IBM Watson to improve intent recognition and conversational interfaces
Personal & side builds
Research Agent
Multi-source AI research agent with knowledge graph, sentiment analysis, token/cost tracker with prompt caching, briefing mode, and model switcher. FastAPI + SSE + D3.
Hover to preview
DocMind
● No document loaded
Drop your document here
Supports PDF, TXT, MD · Max 2 MB
Conversation
💬
Upload a document above,
then ask anything about it.
then ask anything about it.
Ask a question about your document...
Ask →
DocMind — Document Q&A with LLM
Upload PDF, TXT, or MD files and ask questions about them. Uses vector store embeddings and LLM to answer with context from the document.
Hover to preview
🤖
ChatBot
● Online
Type a message...
➤
ChatBot — Conversational AI Interface
Clean browser-based chatbot UI with typing indicator and real-time message rendering. Built with vanilla JS, no dependencies.
Hover to preview
{
"type": "agent_response",
"status": "completed",
"content": [
"text","chart","refs"
]
}
VISUALIZE →
Visual JSON — Agent Response Viewer
Zero-dependency browser tool that parses agent JSON responses and renders them as readable UI — markdown, charts, tables, and reasoning blocks.
Hover to preview
Daily Learnings — Micro Journal
A line-a-day journaling tool with streak tracking and AI-generated monthly summaries. Built with Hatchable + Claude Code MCP integration.
Hover to preview
✓
Read 20 mins
🔥 14 day streak
✓
Exercise
🔥 7 day streak
Ship side project
No streak yet
Habit Tracker with Streaks
Daily habit tracker with streak logic, 7-day week row, and localStorage persistence. Spec-driven from a full product doc.
Hover to preview
ADD R1, R2, R3 — executing
FETCH
DECODE
EXEC
WB
ADD
Fe
De
Ex
Wb
SUB
Fe
De
Ex
Wb
LDR
Fe
De
Ex
CPU Pipeline Visualizer
Interactive walkthrough of the 4-stage CPU pipeline with pipelined view and data hazard simulation.
Background
Education
B.E. in Electronics & Communication Engineering
Adichunchanagiri Institute of Technology, India
Sep 2013 — June 2017
Thought leadership
Technical writing
When One RAG Pipeline Isn't Enough
6 minA beginner's guide to Multi-RAG Systems — when and why to chain multiple retrieval pipelines.
→
Getting Started with LangChain ⛓️
4 minPractical intro to LangChain — chains, prompts, and connecting LLMs to real data sources.
→
Understanding Instruction Authority in LLMs
2 minHow LLMs prioritize system prompts vs user instructions — and why it matters for building reliable AI apps.
→
LLMs Explained Like You're New to AI
3 minPlain-language explainer on how large language models work — no PhD required.
→
Learning by Experimentation
3 minWhy hands-on tinkering beats reading docs — a reflection on building to learn.
→
How to Start Learning AI?
2 minAn honest beginner's roadmap — what to focus on first without getting overwhelmed.
→
AI Basics Everyone Should Know
1 minWhat AI actually is, in plain English — no background knowledge needed.
→
View all posts on Hashnode →7 articles · Notes from My AI Exploration