Senior AI engineering / Pacific Northwest
Senior engineering with agentic leverage. We build the thing that pays back twice.
Fixed-fee discovery, fixed-fee build, your code at the end. For owner-operators ready to fix the thing that’s wasted half a person every week since 2017, and for founders who need senior engineering — including the half their enterprise buyer evaluates on — without giving up cap-table room.
Selected outcomes
The receipts.
Analyst capacity returned to a 300-person product org via a RAG / LLM-SQL onboarding system with deterministic validation and human-in-loop merge.
U.S. patents (8,112,379 · 8,954,370) on Microsoft Policy Processor, shipped in Windows Server, Configuration Manager, and Intune.
Annual engineering investment owned across 39 engineers and four platforms at a public travel company.
Reduction in credit-card fraud losses after re-architecting a monolithic detection service into a resilient real-time microservices platform.
SIEM scale; 40% reduction in IAM operations overhead; 85% reduction in signal-onboarding effort. Multiple platforms, same idea — leverage over headcount.
Software factory in production since February 2026. One senior engineer ships at team-scale timelines with SDLC discipline intact.
These are the receipts. Engagement details under NDA; happy to talk specifics on a call.
How we ship
A multi-agent software factory, run by a senior engineer.
Most shops bill you for the time it takes a team to build the thing. We built a multi-agent software factory — spec-driven, validation-gated, human-approved at every merge — that lets one senior engineer ship at team scale. You get team-quality work on solo-priced timelines, with the SDLC discipline still in place.
That’s the agentic-leverage thesis applied to ourselves first. It’s also why our prices are what they are.
Claude CLI · MCP · SQLite context · Python orchestration · deterministic validation · human approval before merge.
What we say no to
Sometimes the build is “no.”
A platform team once asked us to roll out an enterprise data-quality tool we’d evaluated and found insufficient for their environment. Instead of installing it and writing it up, we built a lightweight, forward-compatible framework that did the job in 200 lines of config, then handed it back to the platform team to incorporate long-term.
That’s the judgment you’re buying. The build comes after the judgment.
Who we work with
Two kinds of customers, named.
The thing that’s wasted half a person every week since 2017.
You’ve watched “AI-powered” get slapped onto every competitor’s homepage. You’ve also watched the thing you actually need — the spreadsheet that runs Tuesday, the clipboard nobody trusts, the handoff between sales and ops that loses a deal a month — stay broken for years.
We build the fix. The boring one. The one that pays back in a quarter and is still running in five years. Sized for businesses that need real engineering and can afford it, where a bespoke build returns multiples of its cost.
For operatorsYou raised on the model. You’ll ship on the system around it.
Auth, governance, integrations, evals, audit logging, the security questionnaire that closes the enterprise deal — that’s the unglamorous half of your wedge and it’s what we build.
Cash is the default. Equity is optional, on terms negotiated per engagement. We will not pressure you on cap-table room.
For foundersWhat an engagement looks like
From first email to clean handoff, in weeks, not quarters.
Discovery
We map the painful thing in your words, write the scope, and price the build. Walk away with a real plan whether or not we keep going.
Build
We ship the software, on the scope and price set in discovery. Weekly demos, no surprise change orders, no hourly billing games.
Handoff
You own the code, the docs, and the runbook. The handoff is the goal, not a sales gate. No lock-in.
Embedded operator
If you want us to keep running it or to act as the senior engineer on the system you depend on, we will. If you want it in-house from day one, take it.
Services
A short menu, priced.
Who’s behind this
Jeremy Nay leads it. The “& Co.” is a real bench.
Twenty-plus years of engineering leadership at Microsoft, SDL, and Expedia Group — including a director seat over 39 engineers and four platforms on an $11.4M annual investment, and two U.S. patents on the Microsoft Policy Processor shipped in Windows Server, Configuration Manager, and Intune. Now running a small shop on his own terms: senior, opinionated, fixed-fee, with a multi-agent factory underneath. The principal is always on the engagement.
McNee & Co. is the customer-facing name for Nay Systems LLC. There’s a short story behind the name.
Selected work
Things we’ve built. Things we’re building.
Portfolio management software, end-to-end.
Portfolio modeling, rebalancing workflows, market analytics, and an AI-assisted engineering workflow under the hood. Concept to usable product, solo principal.
Read the case study → Local retail operationsFull-stack ops platform for an owner-operator.
Inventory, scheduling, and admin workflows integrated into one practical system the owner actually uses. Replaced four disconnected tools.
Read the case study → AI-native consumer productMulti-agent career coaching platform.
Sourcing, application, and coaching agents wired into a single product for role discovery, resume targeting, application strategy, and job-seeker guidance. Multi-agent system in production.
Read the case study →