Recruiters & Hiring Leaders
Evaluate for a principal or director role
Hiring overview with LinkedIn recommendations, proof metrics, live platforms, and a forwardable executive brief — built for principal and director screens.
Open hiring page
Open to opportunities
Principal AI architect — production agent systems you can inspect before we talk.
I design and ship governed agent platforms — orchestration, policy gates, enterprise RAG, and eval discipline. Panels start on a 5-spine review path; the full catalog and open repos are there when you want depth.
19 years at Google, Kaiser, Volvo, and Lucid · multi-agent platforms in production · reference stack you can fork, not slide-deck AI.
After 19 years building systems under real operational constraints, I treat agents as governed distributed systems—not prompts with tools.
5
Spine review path (govern · orchestrate · RAG · publish · ADRs)
18
Catalogued ecosystem repos
29
Documented ADRs
19+
Years shipping production systems
For engineering panels → 5-spine review path (4 live + ADR hub) · Full catalog (12 catalog platforms)
Enterprise delivery + open-source proof
19+
Years in enterprise systems
10→2
Agent ops staffing reduction
Multi-$M
Revenue & savings delivered
12
Production platforms shipped
20+
Engineers led
5
Governed AI platforms in stack
Latest shipped
Build log — what changed recently
PortfolioJul 7, 2026
DomainForge: RAG + QLoRA support triage pipeline
SOP corpus RAG, Bitext ChatML SFT, S0–S2 eval compare, TRL training CLI, and live UI.
PortfolioJul 6, 2026
Portfolio: platform-first proof blocks + technical review path
Decision/signal/limitation on every platform card, 5-minute panel checklist, hire audience toggle.
PortfolioJul 1, 2026
Enterprise RAG: live PDF upload + three retrieval strategies
Access-aware hybrid RAG with side-by-side strategy comparison on the live platform.
Start here
Engineers & AI Builders
Follow the architecture writing
Weekly Substack essays, GitHub reference implementations, and portfolio deep dives on agents, RAG, and production AI.
Subscribe on SubstackTechnical Reviewers
Inspect portfolio & live platforms
One page: architecture diagrams, 17 catalog products, open-source repos, and enterprise case studies — no tab hunting.
View portfolio5-spine review path
Govern the agent before it changes the business.
Start with AegisAI (governance), then orchestration, access-aware RAG, and a governed publish path. Labs and curriculum patterns stay in the full catalog — not the panel path.
Canonical case study
AegisAI
Monitor → Govern → Remediate — a runtime control plane for tool authorization, policy violations, HITL approvals, signed audit, and agent registry lifecycle.
Reviewer evidence
Control plane over agent builder — enterprises need governance across fleets, not another chat UI.
Website deploy tools forced through approval_required policy on live platform.
Registry defaults to in-memory on free tier; Postgres path documented for production.
Supporting proof · Multi-agent OS
Venkat AI Platform
LangGraph stateful graphs over linear chains — enterprise workflows need checkpoints and HITL.
Inspect supporting case studySupporting proof · Knowledge layer
Enterprise RAG Platform
Authorization before ranking — vector DB is implementation; access control is architecture.
Inspect supporting case studySupporting proof · Content automation
AI Content Factory
AegisAI gateway blocks publish until policy allows — side effects never bypass governance.
Inspect supporting case studyFlagship thesis
From Multi-Agent OS to Agent Governance
Why VAP and AegisAI are complementary layers — not competing products.
Eval discipline
golden-eval-registry — regression gates wired into real CI
Six suite kinds (RAG answer, harness QA, repo fix, mission gate, graph HITL, brief gate). Enterprise RAG and AegisLoop fail builds on regression — not just validate fixtures. Running one suite for the first time found and fixed a real bug in its own fixture.
Writing
A connected body of work—not isolated posts.
18 Substack essays since April 2026, moving from agent patterns to the enterprise control plane—and linking each thesis to inspectable implementation proof.
SubstackJul 27, 2026
The Governed Agentic Supply Chain
Supervised autonomy—not autopilot—is the production architecture enterprises need
SubstackJul 23, 2026
Federated RAG Has a Routing-Integrity Problem
Why keeping enterprise data local does not guarantee trustworthy retrieval
SubstackJul 21, 2026
The Agent Trajectory Optimization Layer
Why enterprise agents need more than traces, dashboards, and token metrics
Get in touch
Hiring for a principal AI architect, or want the technical review first?
Read the hiring overview, browse the full portfolio, or subscribe for weekly architecture essays.