AI-Native Engineer · Technical Architect
11 years turning 0→1 ideas into production systems. I embed with teams as a Forward Deployed Engineer — building governed multi-agent AI, RAG copilots over MCP, and predictive engines, and owning them end to end.
Foundation models (OpenAI, Claude, Hugging Face, LLaMA) applied to governed asset agents, MCP-connected RAG copilots, and LangGraph multi-agent orchestration — with fine-tuning, guardrails, evaluation, and human-in-the-loop controls.
Serverless ETL, medallion architectures, and delta pipelines over millions of records — with governance and validation built in.
Protocol engineering, decentralized identity, and scalable backends in Go and Python — from consensus layers to serverless platforms.
Embedded with clients to ship Generative-AI products from discovery to production: governed multi-agent asset agents, a RAG Copilot over secured MCP, a predictive engine for healthcare HVAC, and a governed IT Support Copilot on Azure for multi-tenant organisations.
Built the AIDLC BRD engine — a LangChain / LangGraph multi-agent system — from ideation to production, led enterprise migration to Salesforce with full governance, and shipped serverless delta loads plus a Databricks medallion architecture.
Led an esports GameFi ecosystem from ideation to production — 400K+ decentralized IDs via Privado ID, L1 development on Avalanche Subnet and Polygon CDK, and a re-architected serverless Tezos launchpad.
Enterprise asset-management agents that forecast failures, analyse operations, and act within policy guardrails — multi-agent, auditable, and governed by Responsible AI and human-in-the-loop controls.
A RAG assistant that plugs into any system through a secured MCP connection — using tool/function calling and semantic search to expose one conversational interface over scattered enterprise tools.
An AI Development Life Cycle (AIDLC) engine orchestrating LangChain / LangGraph multi-agent workflows that take business requirements from ideation to production, with LLM evaluation and governed medallion data.
A policy-driven, governed AI copilot for multi-tenant organisations — a governed agentic workflow with three-tier RAG (hybrid pgvector + BM25, RRF fusion) on Azure, grounded by content filtering and groundedness evaluation, with structured auto-escalation to IT.
Let's build
I embed with teams as a Forward Deployed Engineer — taking agentic systems and data platforms from discovery to production. Let's talk about where you're headed.