AI Engineer focused on production-grade LLM applications, agentic systems, RAG pipelines and cloud-native AI platforms.
I build practical AI systems that connect business problems to reliable software: APIs, agent workflows, retrieval pipelines, evaluation layers, observability, deployment and operational guardrails.
- Python, FastAPI, Pydantic
- LangGraph, LangChain, LlamaIndex
- RAG, Agents, Tool Calling, MCP
- PostgreSQL, Redis, Qdrant
- Docker, Kubernetes, AWS, Azure
- OpenTelemetry, Prometheus, Grafana, Langfuse
- Evaluation, Guardrails, Cost and Latency Monitoring
Platform for building, evaluating and observing LLM agent workflows. Includes FastAPI contracts, workflow orchestration, evaluation hooks, trace metadata, Docker Compose and tests.
Hybrid retrieval system with citations, reranking, Recall@K / MRR evaluation and an API surface for enterprise knowledge-base search.
Agentic RAG system in Python for intelligent document search, retrieval workflows and AI orchestration.
Applied data and AI monitoring system for public-health reporting, deterministic analytics and automated reports.
Parameter-efficient fine-tuning project with LoRA, T5, ROUGE evaluation, ablation, demo, tests, CI and Docker.
Fullstack technology consulting platform built with TypeScript, React, Vite and Tailwind CSS.
- Telecom AI Agent: multi-agent customer support workflow inspired by telecom operations.
- LLM Evaluation Lab: framework for groundedness, latency, cost, refusal and hallucination risk.
- Prompt Security Gateway: protection layer for prompt injection, PII detection and audit logs.
I am building toward Forward Deployed AI Engineering: translating complex business problems into reliable, observable and scalable AI systems.