Backend & Applied AI Engineer
I build backend systems and infrastructure around LLM applications — reliable ingestion pipelines, RAG systems, agent observability, evaluation, and developer tooling.
- LexGuard — durable document ingestion and retrieval using Spring Boot, FastAPI, PostgreSQL/pgvector, RabbitMQ, transactional outbox, recovery workflows, and observability.
- TraceLab — LLM agent observability and evaluation platform with a published Python SDK.
- Enterprise Agentic RAG — conditional retrieval, reranking, guardrails, and evaluation with LangGraph and Qdrant.
Python · Java · FastAPI · Spring Boot · PostgreSQL · RabbitMQ · Redis · Docker · LangGraph · Qdrant · Next.js
Projects & demos: sverma.online
Email: svermaengineer@gmail.com
LinkedIn: linkedin.com/in/sonu-verma28
