⚔️ The Code Forge ⚔️
Light
Dark
System
AI & Machine Learning

Building Production AI Systems

From enterprise document intelligence at Hyland to multi-provider LLM infrastructure and RAG-powered chatbots — AI is the sharpest edge in my toolkit.

5+LLM providers supported
3Production AI systems
M+Documents processed
4+Years AI development

AI Projects

Production systems built with modern LLM tooling across enterprise and personal contexts.

🏭

Hyland Automate — Agentic Document Intelligence

Enterprise · Hyland Software

Building agentic AI features inside Hyland Automate, a content services platform used by Fortune 500 companies. Users create workflows and applications through natural language prompts; the agentic layer interprets intent and generates low-code configurations automatically.

  • Natural-language-to-workflow generation via agentic AI
  • Document categorization, extraction, and classification at enterprise scale
  • Confidence scoring and human-in-the-loop patterns for edge cases
  • Intelligent search across enterprise document repositories
  • Multi-step reasoning workflows with tool-use / function calling
Agentic AI.NET CoreAngularLLM IntegrationDocument Intelligence
🔀

LLM Proxy Server — Multi-Provider Architecture

Personal Project · Production

A production-grade .NET Core proxy that provides a unified API surface over five LLM providers. Routes requests to the most cost-effective model for each task with real-time cost tracking and quota management.

  • Unified API over OpenAI, Anthropic, Azure AI Foundry, AWS Bedrock, and Groq
  • Real-time cost tracking and spend visibility per request
  • Intelligent routing — selects optimal model based on task type and cost
  • Rate limiting, quota management, and provider-level fallback
  • Request/response logging with full audit trail
OpenAIAnthropicAWS BedrockAzure AI FoundryGroq.NET Core
🤖

Portfolio AI Assistant

Personal Project · Live on this site

The AI chatbot powering this portfolio. Built with Next.js and the Vercel AI SDK, it uses a RAG pipeline backed by a Python/FastAPI service with ChromaDB for vector search. Supports multiple conversation modes — technical deep-dive vs. quick hiring overview.

  • RAG pipeline: personal knowledge base + vector search via ChromaDB
  • Multi-model support: switches between Claude and GPT-4 at runtime
  • Conversation mode switching for different audience contexts
  • Streaming responses with typing indicators
  • Source citations surfaced inline in chat responses
Next.jsVercel AI SDKChromaDBPythonFastAPIRAG
💬

Try the AI Assistant

Ask me anything about my AI work, architecture decisions, or how I could help with your project. The assistant uses RAG over my actual experience and project history.

Discuss an AI Project

⌘Kto open from anywhere

AI Tech Stack

LLM providers, vector databases, frameworks, and tooling I work with regularly.

🧠OpenAILLM Provider
🔮AnthropicLLM Provider
☁️AWS BedrockLLM Provider
🏢Azure AI FoundryLLM Provider
⚡GroqLLM Provider
🔗Vercel AI SDKFramework
🗄️ChromaDBVector DB
🐍PythonBackend
🏗️.NET CoreBackend
📊pgvectorVector DB
🔧LangChainPatterns
✍️Prompt EngineeringSkill
← View Full Portfolio