Sumit V.Private profile: Upwork shows its earnings only to signed-in members, so we check it every few weeks, not daily. MRR here is a rough estimate (Upwork's six-month earnings ÷ months active). Last read Nov 5, 2026.

Healthcare AI Engineer + Physician (MBBS, MD) |  HL7/FHIR, EHR/EMR Integration, HIPAA I Build Production LLM Pipelines, RAG & Agents (MCP, LangGraph) | Python, Go, AWS 👁‍🗨 Overview I build and ship production AI systems myself, end to end. No handoffs: I design the architecture, write the code and stay on it until it is deployed, monitored and in use. I bring 10+ years of production engineering inside a 25-year career across clinical medicine and enterprise technology, an MBBS and an MD and years in the place AI gets hardest, regulated healthcare data. ✅ What I Build: • Healthcare AI & Interoperability: HL7/FHIR/C-CDA integration, EHR data pipelines, clinical coding (ICD-10, SNOMED-CT), HIPAA systems, PHI redaction, zero-retention model configurations • AI Agents & Tooling: MCP (Model Context Protocol) servers and tool integrations, LangGraph and LangChain orchestration, function and tool calling, agentic workflow automation • LLM Applications: RAG pipelines, document AI, structured data extraction, classification and summarization, semantic search • Full-Stack AI Products: Python and Go backends, React frontends, Docker and Kubernetes, CI/CD on AWS 🎯 Hands-On Builds: • GoDaddy: built the production LLM pipeline for clinical and operational document intelligence, with RAG grounding, schema-constrained output, confidence thresholds routing uncertain cases to human review, and PHI redaction before model calls • Fulcrum Digital: as lead engineer, built the EHR integration layer mapping HL7 v2.x, FHIR R4 and C-CDA in both directions across Epic, Athena and Cerner, with ICD-10/SNOMED-CT coding and HIPAA audit trails • Juvare: built clinical rules engine modules and FHIR/HL7 exchange driving decision support and alerts for hospitals, EMS and public health agencies, in service through COVID • DISQO: owned identity resolution and deduplication across partner feeds, with fuzzy matching and candidate scoring on high-throughput Go and PHP services • SafeGuard: led legacy modernization, breaking PHP monoliths into microservices for real-time contract exchange across a dealer network under insurance compliance rules 💼 Industry Expertise: • Healthcare: EHR/EMR integration, clinical decision support, public health and EMS systems, HIPAA and BAA work • Insurance & Claims: claims and payer data exchange, warranty and finance administration, regulatory workflow automation • Consumer Data & SaaS Commerce: high-throughput ingestion, identity resolution, multi-tenant integrations 🔧 Technical Stack: AI/LLM: Claude on AWS Bedrock, OpenAI, Azure OpenAI; RAG, vector search (pgvector, OpenSearch), prompt engineering, structured output, self-hosted models for private data Healthcare Standards: HL7 v2.x, FHIR R4, C-CDA, ICD-10, SNOMED-CT, RxNorm; Epic, Athena, Cerner, NextGen Agents: MCP, LangGraph, LangChain, tool and function calling, agentic workflows Evals & Guardrails: labeled eval sets, per-release scoring, precision and recall tracking, confidence thresholds, human review routing, PHI/PII redaction ML & NLP: classical NLP, OCR, document processing, clinical coding automation, rules engines, matching, deduplication, scoring Languages & Frontend: Python, Go, PHP (Laravel, Symfony), TypeScript, Node.js, SQL, React Cloud & Data: AWS (Lambda, ECS, RDS, S3, CloudWatch), Docker, Kubernetes, PostgreSQL, MySQL, Redis, RabbitMQ, Datadog, CI/CD Architecture & Security: REST, GraphQL, webhooks, microservices, domain-driven design, TDD; OAuth 2.0, SAML, mutual TLS, RBAC, encryption, audit logging 📊 Impact: • Document automation that moved staff off routine processing and onto exceptions, with accuracy tracked per document type • Patient records synchronized in both directions across four EHR vendors behind a single data model • Decision support in service at hospitals and public health agencies through the COVID surge 🎓 Background: • MBBS and MD. Physician before engineer, about 25 years across clinical medicine and enterprise technology • 10+ years shipping production software across five employers, from PHP modernization to Go microservices to GenAI :star2: What Sets Me Apart: I am not only Software Engineer but Physician. When your clinical team describes a workflow I follow both the medicine and the engineering, and I know what an auditor will ask for a year from now. Most AI projects fail in the operational half, not at the model: evaluation, compliance, observability and what happens when the model is unsure. That half is what I have done longest, so compliance, evals and monitoring are built in from day one. 🌟 Working With Me: Based in Atlanta on Eastern time, available across client timezones. Have an AI feature or integration that needs building? Send me the details and I will reply with how I would build it.