Experience
Nine years of Java and distributed systems, then three years of building production AI on top of them.
Roles
Production AI (2023–now)
-
Senior Software Engineer
Enterprise Multi-Agent AI Platform
- Own architecture and design decisions for an end-to-end enterprise AI platform on AWS (backend services, data pipelines, APIs and the agent layer) that automates multi-step document work business teams previously did by hand; cut processing time by ~60% and model running cost by ~55%.
- Work directly with client stakeholders, from engineers to executives, to break down workflows, agree on what “good” looks like, and turn it into a technical design and delivery plan.
- Built specialized agents coordinated by a supervisor over A2A messaging and a shared MCP layer, so workflows no single model handled reliably now run start to finish without a human in the loop.
- Added guardrails for input and output safety, PII protection and prompt-injection defense, backed by automated evaluation (RAGAS) and regression suites that keep quality consistent across releases.
Stack: Python, FastAPI, AWS Bedrock (Agents, Knowledge Bases, Guardrails), SageMaker, Lambda, Step Functions, EKS, Aurora PostgreSQL; LangChain, LangGraph, CrewAI; GPT-4o, Claude; pgVector, FAISS; Terraform
Document Workflow Orchestration
- Built a pipeline that turns unstructured client documents (briefs, emails, transcripts) into schema-aligned JSON, routing low-confidence fields to a human reviewer instead of guessing; held 90%+ field accuracy in production.
- Implemented structured extraction with GPT-4o and Pydantic validation (required-field checks, time-format normalization, confidence flags), and reduced fabricated fields by enforcing explicit “unknown / needs confirmation” behavior.
- Built FastAPI services for client-side integration with client engineers, covering contracts, error handling and throughput.
Stack: Python, FastAPI, OpenAI GPT-4o, Azure OpenAI, LangChain, Pydantic; AWS Lambda, EKS; Docker, Kubernetes
Job Semantic Search Platform
- Built the embeddings ingestion pipeline and semantic search APIs over Aurora PostgreSQL with pgVector, combining vector similarity with metadata filters and tuning relevance and query performance to interactive response times.
- Added monitoring for ingestion failures, retrieval latency and relevance debugging, and set team standards for RAG patterns, structured outputs, evaluation and observability.
Stack: Python, Azure OpenAI embeddings, Aurora PostgreSQL + pgVector, AWS EKS, S3; REST APIs, Kubernetes, CloudWatch
Travel Planning Copilot & Virtual Concierge
- Delivered a customer-facing travel assistant with deterministic itinerary generation using structured templates and schema-driven LLM output, integrated with the Google Places API.
- Built and deployed the React chat front end over the RAG pipeline and FastAPI backend, including loading, error and empty states, so the client could use it directly.
- Shipped confidence-based clarification, safe refusal and escalation paths so the assistant asks instead of giving confident wrong answers; hardened the APIs with retries, timeouts and graceful degradation, with latency and fallback metrics and regression prompt suites.
Stack: Python, FastAPI, React, TypeScript; OpenAI GPT, Azure OpenAI, LangChain, Pinecone, pgVector, Pydantic; AWS Lambda, EKS; CloudWatch, LangSmith
Java and distributed systems (2014–2023)
Java Developer, Enterprise Microservices Platform · TIAA
Built Spring Boot REST APIs and microservices, owned deliverables end to end with zero production incidents, and drove SonarQube adoption across the team.
Java Developer, Financial Change Management · Wells Fargo
Delivered Spring Boot REST services and microservices as an individual contributor working directly with US counterparts, resolving upstream and downstream issues within SLA.
Java Developer, Product Data Management · Client: Apple
Built Spring MVC and Spring Boot applications across the full SDLC, with Java/J2EE services over REST and SOAP and a Hibernate data-access layer.
Java Developer, SOA Integration & Healthcare Systems · Napier
Delivered SOA integrations and REST services for Tier-1 telecom and healthcare clients, including the AAASC authentication and authorization API for Telstra and core modules of the Napier Hospital Information System.
Stack (2014–2023): Java 8, Spring Boot, Spring MVC, Hibernate/JPA, JAX-RS, REST, SOAP, Kafka, Oracle, PostgreSQL, JUnit, Mockito, Maven, Jenkins, SonarQube
Education
Master of Computer Applications (MCA) and Bachelor of Science (B.Sc.), Acharya Nagarjuna University, India
Skills
- Front end
- React, TypeScript, JavaScript, Streamlit
- Backend and distributed systems
- Java 8/11, Spring Boot, Spring MVC, Python, FastAPI, REST, GraphQL, SOAP, microservices, Kafka, OAuth 2.0, mTLS
- Data
- Aurora PostgreSQL, Oracle, DynamoDB, pgVector, FAISS, Pinecone, ETL and embedding pipelines
- AI
- Claude, GPT-4o, AWS Bedrock, Azure OpenAI, LangChain, LangGraph, CrewAI, RAG, multi-agent orchestration, MCP, guardrails, RAGAS evaluation
- Cloud and DevOps
- AWS (Lambda, Step Functions, API Gateway, EKS, S3, IAM/KMS), Docker, Kubernetes, Terraform, GitHub Actions, Jenkins, CloudWatch
Last updated: October 2026