Venkatesh Guttikonda / AI Systems Engineer
Reliable AI systems, without demo noise.
I build agent runtimes, retrieval systems, eval harnesses, and applied LLM products where behavior is observable, bounded, and useful.
Agent StackLangGraph · LangChain · Pydantic AI · MCP · Semantic Kernel · Claude Agent SDK · ReAct · Tool Calling
RAG & RetrievalQdrant · Pinecone · FAISS · FastEmbed · LlamaIndex · GraphRAG · Hybrid Search · Reranking
APIs & InfraFastAPI · Next.js · Python · TypeScript · Postgres · SQLite · Docker · Kubernetes · Azure · GCP
Evals & ObservabilityRAGAS · DeepEval · Phoenix · LangSmith · OpenTelemetry
Featured AI systems
A focused set of systems that show what I have built and the engineering problems I can handle in real teams.
The story track
These are the skills the projects are meant to prove: building reliable agent workflows, grounding retrieval, measuring quality, and turning AI into usable software.
Agent runtimesPlanning, tool use, validation, orchestration, and run artifacts.
RAG and contextRetrieval quality, citations, token budgets, context packets, and grounding.
Evals and observabilityMetrics, traces, quality checks, regression gates, and audit surfaces.
Applied AI productsUsable apps that wrap AI behavior in clear workflows and guardrails.
Developer systemsTools that make AI engineering repeatable across repos and teams.
Project index
A broader view of the systems, tools, and applied AI products I have built. Filter by the kind of engineering signal you want to inspect.