def prasad():
return {
"currently": "building agents that think, not just respond",
"background": "backend-leaning full-stack dev, 4+ yrs in production systems",
"past life": ["C3 AI → GPT-4o chat interface, agent tools, dynamic SQL",
"Vodafone → Kafka pipelines, 12k+ events/day, financial microservices",
"AutoPCB → sole engineer, built the full platform from scratch"],
"exploring": "agentic AI systems + distributed backends that power them",
"fun fact": "i once debugged a race condition at 2am and felt nothing",
}Cancer Clinical Trials Agent — a multi-agent system that matches patients to relevant clinical trials.
patient query → intake agent → trial retriever (pgvector) → eligibility checker → ranked matches
LangGraph for orchestration. Temporal for durable workflows. Kafka for async event streaming. pgvector for semantic search. OpenTelemetry for full observability across agents.
the kind of system where "it works" means someone might actually find the right trial.
| area | tools |
|---|---|
| backend | Python, Java, FastAPI, Django, Spring Boot, Node.js, REST, GraphQL |
| frontend | React, TypeScript, Next.js |
| AI/agents | LangGraph, LangChain, OpenAI APIs, RAG, pgvector |
| messaging | Kafka, RabbitMQ |
| databases | PostgreSQL, MySQL, MongoDB, Redis |
| infra | Docker, AWS, GitHub Actions, k8s |
open to backend, full-stack & AI engineering roles. if you're building something interesting, let's talk.