99.9%
Production uptime with 100k+ daily API requests.
Senior Applied ML | GenAI | MLOps
I help teams ship reliable AI products faster by combining graph-aware LLM workflows, robust ML engineering, and cloud-native MLOps discipline. My focus is reducing failure risk while increasing measurable delivery impact.
Production uptime with 100k+ daily API requests.
Pipeline acceleration through workflow hardening.
F1 score increase in entity extraction systems.
OCR extraction accuracy across 5M+ lines.
Graph-memory storytelling engine that achieved 61% fewer contradictions in paired evaluation using Neo4j, dual RAG retrieval, and pre-generation guardrails.
View RepositoryNeo4j + FastAPI graph-backed retrieval with a multi-stage Cypher pipeline for grounded RAG.
View RepositoryDistributed RL platform with Ray and PyTorch PPO using fault-tolerant workers and async rollouts.
View RepositoryAI-enabled automation pipeline to transform instructional PDFs into reusable presentation assets.
View RepositoryPractical applied ML and data science implementations showcasing end-to-end delivery skills.
View RepositoryMCP-focused prototype with LEPOR-driven evaluation concepts for practical AI workflow experimentation.
View RepositoryLangChain, LangGraph, structured outputs, workflow orchestration, and contradiction guard patterns.
Neo4j 5.x, Cypher, Graph RAG plus Vector RAG, ChromaDB, and causal memory retrieval pipelines.
PyTorch, Transformers, OpenCV OCR, NER pipelines, structured extraction, and benchmark harness design.
Python, FastAPI, Flask, Streamlit, Kubernetes, Docker, OpenTelemetry, AWS, Azure, GCP, Databricks.