The AI sub you bring on retainer.
Built from inside the work, not above it. Most AI doesn't fit the business it's sold to — because it's sold from above the work, not built inside it. AI Subcontracting is the install-and-optimize practice for operators already mid-journey with AI: come in, see your operation, identify what's working and what isn't, and build the layer custom to how you actually run.
1–4 wk audit · Monthly retainer · Owned by you
Your AI Sub. On call. On site. On the work.
Inside your operation, not above it. I come in, see what you're already running with AI, name the pain points your team has been working around, and build the capability that fits how you actually work.
Two phases. One relationship.
Start with the audit. If we keep going, the retainer rhythm picks up where the audit leaves off — same person, same context, just a different cadence.
Setup Audit
Site visit, ecosystem walk, workflow audit. We map what you're already running with AI, where work is still manual, and where the highest-leverage gaps live. The output is a named-and-prioritized list of what we'll build first.
Ongoing Retainer
I'm your AI Sub on retainer. Monthly working session, async availability, one capability install per cadence cycle. Workflow optimization, automation builds, prompt tuning, capability transfer to your team. Month-to-month — you stay in control.
The first question isn't “should we use AI?” — it's what shape does this gap need?
Six shapes cover almost every operator-side AI install. Each one fits a different kind of work. The audit names which shapes belong inside your operation, and which don't.
A thinking partner
Sits beside the operator for work that lives in their head. Right when the gap is "I'm doing this alone." Wrong when the work needs to leave your head.
Single-purpose, on demand
Deliberately invoked for a specific job. Right when the task is well-defined. Wrong when the work flows continuously.
Runs over substrate
Operates autonomously across data you already keep. Right when the substrate is real and stable. Wrong when there's nothing under it.
Connects what didn't talk
Glue between systems that never spoke. Right when integration is the bottleneck. Wrong when the systems shouldn't be coupled.
A new face on the work
Customer-facing or team-facing interface. Right when the audience needs a different view. Wrong when the underlying work isn't ready.
Restructures how work flows
Reshapes the sequence itself. Right when the process is the problem. Wrong when the process is already correct.
Most of your operation lives in your head.
The workflows. The handoffs. The decisions waiting on context nobody else has. Bringing on an AI Sub means starting where that mental picture meets the work — and turning the parts that should be automatic into something that runs without you.
- 01Map the workflows that currently live only in your head or your team lead's head.
- 02Identify where work is still manual that doesn't need to be — quoting math, scheduling reads, intake routing, follow-ups.
- 03Build the automations that match how your team actually works (not how a SaaS vendor wishes you did).
- 04Transfer the capability so your team can extend it after I'm gone — not a vendor lock-in.
- 05Tune what you're already running — ChatGPT prompts, vendor tools, Make scenarios — so the pieces start working as a system.
You own what gets installed.
We don't host you, rent it back to you, or sit between you and your customer list. Installed once, owned forever, walk-away-functional.
Your data, your seat
Everything lives in your accounts, your tools, your stack. No hostage data. No vendor dependency we created on purpose.
Transferable to your team
Capability gets documented and walked through. Your team learns what was built, why, and how to extend it.
Month-to-month
No annual contracts. If the retainer stops earning its keep, you stop. The audit stays yours regardless.
AI Subcontracting is often the first door.
When the work surfaces a capability that needs a dedicated specialist agent — vendor intelligence, plant lookup, equipment dispatch, spec sourcing — those Specialists deploy as separate scope. When the work surfaces a multi-trade build, that becomes a Custom Install. The retainer keeps running underneath.
The retainer is the relationship. Specialists and custom builds deploy as scope when the work asks for them.
Owner-operator commercial installers in the $500K–$25M band.
- →You're already running AI in pieces and want it to run as a system.
- →You'd rather bring on a sub than hire a full-time AI person.
- →You want capability transferred to your team — not a vendor relationship you can't exit.
- →You think in subcontractor scope, maintenance contracts, and trade-specific expertise.
- →You want the AI Sub on call when something breaks or a new idea lands.