United States | AI Engineer | Agents · MCP · LLM Systems

Building production-grade AI agents, MCP tooling, and high-throughput LLM inference pipelines.

• 🛟 throughline — Durable execution for AI agents: crash-safe checkpoints, exactly-once effects, durable human-in-the-loop, and replay testing (BYO-LLM)
• 🔭 researchscout — Ingests new AI/ML papers and their signals (citations, code adoption, buzz), then scores, clusters, and summarizes them to separate breakthrough from noise
• 🏗️ agent-scaffold — Turns markdown agent specifications into complete, runnable agent projects in your language of choice

• _{🔭 researchscout — Scores, clusters & summarizes new AI/ML papers to surface what's worth reading right now}
• _{🍽️ mise — Event-driven intelligence layer for restaurant reservation platforms}
• _{🌊 streamcontext — Semantic gateway between Kafka and AI agents; expose real-time data to LLMs via MCP}
• _{🏗️ agent-scaffold — Turns markdown agent specs into complete, runnable agent projects}
• _{🚀 agent-deployments — Production-shaped AI agent prototypes with auth, tracing, rate limits, CI & evals}
• _{🧪 MCP-lab — Test harness stress-testing the Model Context Protocol's security boundaries & spec conformance}
• _{📐 agent-blueprints — Blueprints, patterns & reference architectures for production AI agent systems}
• _{🧱 Structure-D — Ingest unstructured docs into validated, schema-constrained structured data}
• _{🎯 Unified-M — Fuses MMM, incrementality & attribution into one marketing-measurement source of truth}
• _{⚡ GoFuncs — Distributed function execution platform for Go}
• _{🖥️ System-Monitor — Configurable real-time system monitoring tool built in Rust}
• _{🛟 throughline — Durable execution for AI agents: crash-safe checkpoints & replay testing}
• _{💊 Drug-Drug-Interaction-Agent — Drug-interaction analysis agent with RAG, clinical APIs & LangGraph}
• _{💼 Zoe-AI — Business agent with semantic retrieval, LoRA-tuned LLMs & guardrails}
• _{💡 ThinkLife — Nonprofit project (migrated to Think Round, Inc.)}
• _{⚖️ RAG-AI-Compliance-Assistant — RAG + FAISS compliance assistant for legal search & risk assessment}
• _{🎲 Mad-Libs-Generator-Game — Random story generator game}
• _{⌨️ AdvancedKeylogger — Memory-resident keylogger with encryption & email exfiltration (research)}

• Building agentic systems - Production-grade AI agents, MCP tooling, and LLM inference pipelines
• Exploring - Agent evaluation, reliability, and multi-agent orchestration
• Rapid prototyping - Turning research into runnable products
• Open source - Sharing reusable patterns and reference designs for building agents

• Implementing LLM Applications Using LangChain — Parts 1–7 — A 7-part deep dive building LLM apps with LangChain: low-level internals, prompting, chatbots & hallucination control, Pydantic-powered tool agents, LangGraph, RAG with Qdrant, and agent evaluation & monitoring. (Mar–Jul 2025)
• The Neural Code: Encoders and Decoders in NLP — How encoder and decoder architectures work under the hood in NLP. (Mar 2025)
• Harnessing CUDA & RAPIDS for Accelerated Compute — Accelerating data and ML workloads on the GPU with CUDA and NVIDIA RAPIDS. (Mar 2025)

• GitHub achievements: Pair Extraordinaire, Pull Shark, Quickdraw, YOLO, Public Sponsor

"Surpass your limits!" - I build tools to push past what I thought was possible, then share them. Currently exploring how agents change the way we build software.