Software Engineer building the distributed, cloud-native backends behind AI applications. MS Computer Science, Rochester Institute of Technology (2026) · AWS Certified Solutions Architect · Rochester, NY, open to relocation.

Portfolio · LinkedIn · chetan.chandane2026@gmail.com

Three years shipping production software in regulated environments: healthcare and telecom, where an outage is a real problem for real people. I work on the infrastructure layer, the part that has to keep working when traffic spikes, a consumer dies mid-batch, or a model returns something unexpected.

Currently focused on event-driven systems (Kafka, Kubernetes) and the retrieval and evaluation pipelines that make LLM features reliable enough to deploy.

Experience

• Software Engineer, Co-op at Excellus BCBS (Sep 2025 to May 2026). Shipped C#/.NET and Angular features across 5 Azure Functions and ASP.NET Core APIs on a HIPAA-regulated platform. Cut CI/CD runtime 35% and broken builds 30%.
• Software Engineer at Vodafone Intelligent Solutions (Oct 2020 to Jul 2023). Built and maintained Java Spring Boot APIs across a 7-service microservices platform. Shipped an event-driven classification pipeline auto-triaging 5,000 tickets/month.

Documents enter through a FastAPI gateway, stream over Kafka to async workers that run embedding and enrichment, and land in PostgreSQL and Qdrant. Built for the failure cases: a transactional outbox for atomic publishes, Redis-backed idempotent consumers, and a dead-letter queue for poison messages. Runs on Kubernetes with autoscaling driven by consumer lag, traced end to end with OpenTelemetry.

Python FastAPI Kafka Kubernetes PostgreSQL Qdrant Redis OpenTelemetry

Question answering over FDA regulatory documents with inline source citations. Hybrid retrieval (BM25 plus dense, RRF fusion) with a cross-encoder reranker, reaching 0.88 answer faithfulness on a 430-question evaluation set. A CI gate blocks any deploy that drops below threshold, so retrieval regressions never reach production silently.

Python FastAPI Qdrant Ragas LangSmith GitHub Actions

A cyclic LangGraph agent using MCP tool-calling for record retrieval, policy search, and scheduling. A two-stage guardrail (deterministic rules plus LLM verification) held 100% emergency recall while cutting false positives 84% across 189 adversarial cases, correctly escalating keyword-free presentations like silent strokes and CO poisoning. Tool-call error rate of 0% across a 640-run evaluation suite.

Python LangGraph MCP ChromaDB PostgreSQL

Extracts macronutrients from food label photos and generates diet plans. React frontend, 8 REST endpoints on API Gateway, Lambda backend, DynamoDB storage, Cognito auth. Fully provisioned as code with 6 reusable Terraform modules across dev and prod, deployed through GitHub Actions.

React AWS Lambda API Gateway DynamoDB Terraform GitHub Actions

Languages Java · Python · C# · TypeScript · SQL Backend Spring Boot · FastAPI · ASP.NET Core · REST APIs · Microservices · Kafka Cloud and Infra AWS (Lambda, API Gateway, DynamoDB, S3, Cognito) · Azure · Kubernetes · Docker · Terraform Data PostgreSQL · MongoDB · Redis · Qdrant Practices CI/CD (GitHub Actions, Azure DevOps) · pytest · Jest · integration testing · OpenTelemetry · Agile AI RAG pipelines · LangGraph · LangChain · MCP · LangSmith · OpenAI and Anthropic APIs

Certifications AWS Certified Solutions Architect, Associate · AWS Certified Cloud Practitioner

• Evaluation pipelines for agentic systems, moving past vibes-based testing
• Chunking and retrieval strategies for production RAG
• MCP server design patterns

Full-time Software Engineer and AI Engineer roles, US-based, available immediately. Most interested in teams shipping AI into regulated domains where reliability is a requirement rather than a nice-to-have.

Best reached at chetan.chandane2026@gmail.com.