The 90-Day Build

Most small firms have AI. Almost none have a system.

We build the system that runs your sales work, train your team to run it, and hand over the keys in 90 days.

Book a Call

The problem

Four things go wrong in the same order.

The firms that call us have all bought AI already. Most of them live with some mix of these four.

  • Tool sprawl

    Each person has a different AI tab open, and none of it connects.

  • Prompt amnesia

    The prompt that worked last month is gone.

  • Quality lottery

    Two people, same task, different output each time.

  • Invisible risk

    You cannot tell what AI drafted or who checked it.

What we build

Four layers, built in this order.

This is the whole system. Each layer is useless on its own and obvious once the one under it exists.

01

Memory

Your firm's facts, rules and terminology, written down once. Every workflow starts from that instead of from whoever happens to be doing the task.

02

Method

Your recurring sales work, turned into repeatable playbooks. The same job runs the same way whoever picks it up.

03

Governance

A short set of rules for what AI may and may not do here. Every client-facing output passes a review gate before it leaves the building.

04

Operations

Work that runs on a schedule without anyone starting it. We version-control every change, so you can see what moved and put it back.

The 90 days

Three phases, and in the last one you take over.

Book a Call
  1. 01

    Days 0 to 30. Foundation.

    We audit the tools and workflows you run today, build the reference library, and set up the review gate. Your first one or two highest-value workflows ship inside this window.

  2. 02

    Days 31 to 60. Breadth.

    We extend across lead handling, quoting, follow-up and pipeline reporting. We route anything high-risk to a person for sign-off.

  3. 03

    Days 61 to 90. Autonomy.

    Scheduled work goes live and runs on its own. We train and certify your internal owner, then hand the system over.

We build nothing before a written scope, and we are not finished until someone on your payroll can run it without us.

You own it

Three ways to buy AI, and only one of them ends.

All three run on much the same software. They differ on who holds the asset when the work is done.

Beside AI

Own it in 90 days.

  • Your team runs it.
  • We stay on call.
  • The system stays yours.

SaaS subscription

Rent forever.

  • The vendor owns the platform.
  • You never own the asset.
  • The bill does not stop.

Black-box agency

Bills forever.

  • The agency keeps the know-how.
  • You stay dependent.
  • The bill does not stop.

Start here

Start with a 30-minute kickoff and audit.

We map how you use AI today and find the sales work worth automating first. If a build is not worth it for you, we say so on the call.

Who to bring

The owner or GM, and whoever runs your systems day to day.

What you leave with

A one-page summary of where AI would pay off first, whether or not you work with us.

How it fits your IT

The questions that come up on the call.

Don't we already have AI tools?

Yes, and that is the problem you walked in with. Tools without a system do not connect to each other and do not remember anything. We build the layer that makes the tools you already pay for start paying off.

Isn't our industry too different for this?

Your industry shapes the build. It does not change the architecture. Memory, method, governance and operations look the same in a court reporting firm and in a claims shop. You own what goes inside them.

Can't we just build this ourselves?

You could. Plenty of firms get the first workflow running on their own. The hard part is the structure underneath it: the rules, the review gate, the version control, the training. That is what makes the second workflow cheap instead of another project.

How hard is this to integrate?

Lighter than you expect. It runs inside tools you already use: Microsoft 365, Google Workspace, Salesforce. There is no new platform for your team to log into.

How does it fit our IT roadmap?

It lives inside systems your IT already controls and already secures. There is no new vendor platform to review, manage or renew.

What happens when the AI models get better?

You ride the upgrade. Your library, your playbooks and your rules are yours and carry forward. A better model makes the same system better.

Who owns the system when the build is done?

You do. You own your data, your outputs, and everything built on them. We license our methods to you for as long as you run it.

What does it cost?

A fixed-scope build, quoted on a call once we both know what we are building, and then a maintenance and support retainer afterwards. There is no subscription.