Generative AI & RAG
Document ingestion, chunking, embeddings, pgvector retrieval, prompt engineering, and source-grounded LLM responses.
AI Engineer
Building practical, secure, and scalable AI products.
I specialize in production-oriented Generative AI, Retrieval-Augmented Generation, and agentic AI workflows using Python, FastAPI, LangGraph, PostgreSQL, pgvector, and Large Language Models.
~3
Years Experience
RAG
Grounded AI Systems
MCP
Tool-Ready Integrations
Current Focus
Generative AI, RAG, agentic workflows, MCP, and enterprise AI services
About
Production-oriented AI engineering, grounded in enterprise data.
I am an AI Engineer with nearly 3 years of professional experience across Data Analytics and Artificial Intelligence. I specialize in designing and developing production-oriented Generative AI applications, Retrieval-Augmented Generation systems, and agentic AI workflows.
My experience includes document-based AI assistants, multi-agent workflows, Model Context Protocol servers, enterprise chat assistants, and AI-powered backend services. I work across document ingestion, chunking, embeddings, vector retrieval, prompt engineering, tool calling, API integration, response grounding, deployment, and performance optimization.
Before moving into AI Engineering, I worked as a Data Analyst building Python- and SQL-based ETL workflows and integrating QuickBooks, Oracle NetSuite, Maconomy, SFTP systems, and external APIs. That background helps me design AI applications with a strong understanding of data quality, business processes, and enterprise systems.
Expertise
Where I add value
Agentic AI & MCP
LangGraph workflows, multi-agent coordination, tool calling, MCP servers, and human-controlled AI automation.
AI Backend Engineering
Python, FastAPI, PostgreSQL, REST APIs, LLM integration, enterprise chat assistants, and AI-powered backend services.
Production Delivery
Linux deployment, Docker, CI/CD, security-aware design, API integration, observability, and performance optimization.
Selected Work
Private work, public capabilities
Open-Rag-MCP
A private knowledge platform for uploading documents, validating retrieval, and exposing grounded AI answers through a web widget, streaming API, and MCP-compatible clients.
- Workspace-scoped document groups, model configs, ingestion, search bench, agents, citations, and integration keys.
- Designed for BYOK model usage, group-scoped access, one-time secret display, and source-grounded answers.
Private Docs
Search Bench
Agents + MCP
Multi-Tenant Training Portal
An enterprise employee-development platform with tenant-isolated workspaces for administrators, HR teams, trainers, and trainees to manage learning from onboarding through measurable progress.
- Role-governed workflows cover employee onboarding, trainer-to-trainee mapping, syllabus review, assignments, assessments, and progress reporting.
- Built with FastAPI, PostgreSQL, Quasar, Vue, and TypeScript, with private media delivery, encrypted integration secrets, and AI-assisted syllabus authoring.
Onboard
Train
Measure
Workflow intelligence assistant
Designed LLM-backed automation for enterprise operations, combining structured business context, prompt workflows, and API-driven actions.
- Built using Python, FastAPI, LangGraph, and Gemini.
- Integrated with REST/JSON services and internal workflow data.
- Packaged with privacy-safe logging and operational monitoring.
Object detection automation
Developed vision pipelines for detection and decision support, tuned for practical inference constraints and maintainable deployment.
- Trained and evaluated YOLO-based models.
- Optimized inference paths for CPU-oriented environments.
- Connected model outputs to downstream operational workflows.
Production ML delivery pipeline
Created repeatable paths for model experimentation, validation, deployment, and monitoring across containerized services.
- Used MLflow for experiment tracking and reproducibility.
- Containerized services with Docker.
- Automated checks and deployments with GitHub Actions.
ETL and API automation
Built custom pipelines that transform operational data and expose reliable service interfaces for AI-enabled business workflows.
- Implemented data extraction, transformation, validation, and loading.
- Delivered FastAPI endpoints for application integration.
- Improved traceability through structured logs and error handling.
Skills
Technical toolkit
A focused stack for building grounded, tool-using AI systems and taking them from enterprise data to production.
AI Engineering
Generative AIRAGAI AgentsLLM IntegrationPrompt EngineeringResponse Grounding
Frameworks & Protocols
PythonFastAPILangGraphMCPREST APIs
Data & Retrieval
PostgreSQLpgvectorEmbeddingsVector RetrievalDocument IngestionChunking
Production & Delivery
Linux DeploymentDockerCI/CDAPI IntegrationObservabilityPerformance Optimization
Enterprise Data Foundation
SQLETLQuickBooksOracle NetSuiteMaconomySFTPExternal APIs