AI software factory · fractional principal engineer · four-time founder
I tighten how your team builds with AI, and my agentic software factory delivers features at a pace a team can't match alone, still tested and secure.
Previously founded and scaled multiple products including Staging Labs (acquired), Crypto-Loot, ArtlyCart, and uPlexa. Additionally, I have worked with teams at Dapper Labs, Integral, TeraGo Networks, and Deloitte on performance-critical systems.
A 10x engineer was always the one who works ten times more efficiently than the average engineer. Now every engineer has AI, and the average has moved with it.
So the bar moved too: you still have to be ten times more effective than an average engineer who has AI. I get there with a software factory engineered around how I work, and every change it makes is reviewed before it reaches you.
| Week | Opened by my software factory | Opened by hand | Total |
|---|---|---|---|
| 2026-W30 | 0 | 4 | 4 |
| 2026-W31 | 0 | 12 | 12 |
| 2026-W32 | 0 | 9 | 9 |
| 2026-W33 | 0 | 14 | 14 |
| 2026-W34 | 0 | 12 | 12 |
| 2026-W35 | 0 | 20 | 20 |
| 2026-W36 | 0 | 14 | 14 |
| 2026-W37 | 7 | 1 | 8 |
| 2026-W38 | 5 | 2 | 7 |
| 2026-W39 | 6 | 9 | 15 |
| 2026-W40 | 23 | 1 | 24 |
| 2026-W41 | 25 | 5 | 30 |
| 2026-W43 | 36 | 5 | 41 |
| 2026-W44 | 55 | 7 | 62 |
| Total | 157 | 115 | 272 |
Merge requests merged in each ISO week, 2026-W30 to 2026-W44 on one infrastructure-heavy platform project, split into those opened by my software factory under my account and those opened by hand. One week of time off with none merged is not drawn.
These figures come from one infrastructure-heavy platform project with complex tickets: integration and infrastructure work, where each change is harder than a typical application change. They are not from all my projects. A merged merge request is a unit of reviewed change, not a measure of effort. Factory-opened is inferred from the factory branch prefix under my account. The project has no history before July 2026.
Source: one infrastructure-heavy platform project, as of 2026-11-04.
Source: the review ledgers of my software factory
I’m usually brought in when the stakes are high: real users, real money, and systems that can’t afford to break.
Managed engineering factory
I integrate my software factory into your repositories and tune it to your architecture and review rules, so you get pull requests that fit how your team works.
667
Tickets merged by my software factory since 2026-08-19, across five projects, mostly its own development
AI forward deployed engineering
I embed in your problem and build the agents, orchestration and MCP integrations it needs.
The result is running software your team owns, not a strategy deck and not a generic chatbot.
Backend and platform engineering
Distributed systems, infrastructure as code, and the platform work underneath them.
76%
Of merge requests merged touch infrastructure as code
Financial systems and blockchain
Market infrastructure, Solana and EVM infrastructure, and smart contracts built for correctness first.
99.7%
Staging Labs: reduction in execution time
Security automation
Agentic red-teaming: autonomous agents probe your system the way an attacker would, and an engineer reviews every finding before it reaches you.
Every finding is checked by an engineer before it reaches you.
Workflow review · where engagements start
How your team ships today, where AI and my factory fit, and the security issues worth fixing first.
One review of how your team ships, a concrete plan, and you decide whether to continue.
Integral Treasuries
87%
87% improvement in Temporal job processing speed across all chains
Staging Labs
99.7%
99.7% reduction in execution time (≈1s to ≈3ms processing)
LaLiga Golazos
$2.7M
Over $2.7M in first season sales volume
ArtlyCart (Generate-NFT.online)
Enabled non-technical creators to launch full NFT projects independently
uPlexa
First Proof-of-Work blockchain optimized for mobile and IoT hardware
Crypto-Loot
~700,000
~700,000 concurrent connections at peak
A system scales as far as the resource its parts share, so I find that resource and measure time spent waiting for it apart from time spent working.
Most failures in a distributed system happen between its parts rather than inside them, so I design for what each part does when another one goes quiet, answers late or reaches for the same state.
An enterprise system is held together by its controls, so I make each merge gate read the evidence it claims to read and give each person and service only the access its current work needs.
When work moves fast, an approved plan is wrong in small factual ways by the time it is built. The fast path that stays correct is a builder who checks each premise against the code, keeps the plan’s intent, and writes down every place the plan was wrong and why.
A person in the loop is a throughput dependency, not only a safeguard. When every hold on landing work waits for that person, an autonomous system stops landing anything and looks healthy while it does, so it has to report what landed and what is waiting on someone.
I have founded products and hired a team of engineers to grow one, and running them meant decisions engineering does not answer: who the product answers to, the smallest version people can try, whether they use it and pay for it, and what a no should trigger.
I help crypto and TradFi teams design and ship production systems that can scale without rewrites, firefighting, or hidden technical debt.
I look at how work moves through your team today, from ticket to production: where AI helps, what my software factory can take on, how to configure it for your codebase, and the security issues I find along the way. You leave with a plan either way. From you: access to your repositories and CI, and a walkthrough of how your team ships.
I set the factory up in your repositories and tune it to your architecture, conventions and review rules. This is when I work most closely with your team, agreeing how work reaches the factory and how its changes come back for review. From you: your conventions, your review rules, and the people who know them.
The factory plans, builds and reviews work from your roadmap, and each change reaches you as a reviewed pull request. You get architecture direction from me. From you: priorities, and sign-off where your approval workflows ask for it.
One review of how your team ships, a concrete plan, and you decide whether to continue.
[email protected]