Record actions, add image and text checks, then refine the logic in a visual editor.
One workflow. Multiple ways to edit.Recording, visual blocks, and code share the same automation format.
Engineering details
Vision-driven control: OpenCV template matching and
cross-platform OCR (EasyOCR on Windows, Apple Vision on macOS) driving
devices over ADB.
An automation IR — one serializable step-tree that a
recorder, a drag-and-drop block editor, and a code view all round-trip
through.
Multi-tenant cloud control plane: a websocket relay,
accounts/quotas/billing, and on-demand cloud Android emulators streamed
to the browser.
Over 100k lines across the platform; thousands of automated tests; CI
across a Python version matrix and a signed Windows installer
pipeline.
Architecture overview: local models and memory, with policy-controlled cloud routing.
Private conversational AI
AI that remembers. Data you control.
A self-hosted conversational engine with durable memory and model routing built around task, cost, and privacy.
Privacy expressed in code.Automated tests check that sensitive text stays within the intended boundary.
Engineering details
Multi-model router that chooses per turn between
local open-weight models and cloud APIs by task, cost, and privacy.
Long-term memory on Postgres + pgvector — hybrid
full-text and vector retrieval with reciprocal-rank fusion and
importance/recency reranking.
Privacy enforced by tests: a suite that guarantees
sensitive text never reaches a cloud provider — a guarantee in code, not
in a policy doc.
~2,400-test regression gate; an optional multi-tenant layer with
auth, metering, payments, and fail-closed safety gates.
The controller’s actual dashboard, replaying bundled sample data from June 15, 2024. No devices connected. View full dashboard
Energy & control systems
Put surplus solar to work.
A deployed controller directs excess solar into EV charging while managing home-battery targets.
Engineering details
Physics-based solar forecasting calibrated against
real meter data, feeding a 72-hour battery trajectory model with a hard
floor it will never cross.
Safety-layered device control: dry-run by default,
advisory mode, rate-limit and stale-data guards, geofencing, and a
persisted decision audit trail.
Deployed via Docker on a cloud host; 100+ tests including a historical
backtest simulator that replays past days.
Business automation
Autonomy with clear limits.
An operations agent computes net profit, makes low-risk improvements, and sends money decisions to a human.
Engineering details
Five-tier fail-closed autonomy: unknown actions
default to "requires a human," and destructive or terms-of-service-risky
actions are hard-blocked, never auto-run.
Hash-chained audit ledger: every action is
SHA-256-linked to the one before it and independently verifiable — if the
system can't prove what it did, it doesn't ship.
Live in production against real marketplace APIs, with a
research-to-decision paper trail behind every rule.
Personal finance
Turn account data into action.
A local-first engine surfaces cash runway, duplicate charges, and subscriptions, with read-only access.
Engineering details
Rules-based insights engine (cash runway, duplicate
charges, subscription audit, and more) that re-runs on every sync and
remembers what you've dismissed.
Diffed automation ledger that classifies each item as
new, continuing, snoozed, or resolved and renders a daily brief.
Read-only data scope, localhost-bound, CSRF-guarded — safe by
construction, not by trust.
Where I can help
A focused architecture review, a difficult integration, or a system built from the ground up.
Agents & automation
Connect AI to real workflows, with clear approvals and audit trails.
LLM & MCP integration
Bring models and tools into your stack with reliable routing and fallbacks.
Retrieval & memory
Find the right information through hybrid search and durable memory.
Private AI
Run on your infrastructure with explicit, tested data boundaries.
How I work
Hands-on, end to end
I write the code and ship working software, then leave your team able to run and extend it.
Verification built in
Tests cover how the system should work, how it can fail, and the boundaries it must respect.
AI-native delivery
AI coding agents work within isolated workstreams, verification gates, and regular audits.
A pilot’s approach to engineering.
Aviation trains you to think about what can go wrong—and how you’ll know before it does.
I’m an airline pilot who builds production software. I’ve designed and shipped AI systems solo, from a private conversational engine to autonomous agents that operate real businesses.
I bring the same discipline to both: explicit limits, clear audit trails, and verification that makes “it works” something you can demonstrate.
What are you trying to build?
Tell me what you’re building, what’s getting in the way, and what constraints matter.
