The foundation
Design the company first.Then automate it.
Every AI improvement we build sits on one blueprint: three layers, three pillars of value, and the memory underneath them. Pulse is the first engine we built on it. It won't be the last.
The blueprint, as a map
Click any part to read what it does and where the machine and the people each fit.
Pulse runs the Leveraged Sales pillar today.
Most AI projects automate the tangle. Faster confusion is still confusion.
In most companies, the person doing the work also reports on it, interprets the numbers and argues strategy, often in the same meeting. Method lives in a few senior heads. Two systems disagree about the same deal, and someone reconciles them by hand every week.
That shape made sense when moving information was expensive. Point AI at it and you get a quicker version of the same mess: more dashboards, more summaries, more meetings about what the summaries mean.
So we start with the structure the automation will run on.
Three layers, each with one job.
Deciding where to go, tracking whether you're getting there, and doing the work customers pay for are three different kinds of work. They interfere when they're mixed, so the blueprint separates them and treats meetings as the wiring between them.
Sets direction
Lead
A mission that names the seam the company sells into, and five filters every real decision has to pass, applied out loud by whoever is deciding.
- The machine
- Surfaces the filters at the point of decision and keeps the record of which one carried each call.
- The human
- Sets the mission and makes the call when two filters pull in different directions.
Measures, manages, mentors
Measure
One scorecard per pillar, nine numbers in all, managed against and used to grow people, not only to score them.
- The machine
- Keeps all nine numbers live from the work itself and flags a divergence while it's still cheap.
- The human
- Decides what a moving number means, acts on it, and explains it to the person who can fix it.
Creates the value
Do
Repeatable value engines with owners and playbooks, and service lines that share one shape, so delivery can be handed off and scaled.
- The machine
- Runs the uncertainty-reducing parts: assembling the deal picture, tracking project state, watching account health.
- The human
- Does the expert work, owns the relationships and the hard conversations, and improves the engine.
Three pillars. Nine numbers. Not ninety.
| Pillar | Its scorecard |
|---|---|
| Leveraged Sales: growth that doesn't need proportional headcount | Pipeline, win rate, deal velocity |
| Transferrable Delivery: work that survives the person who invented it | Utilization, customer satisfaction, on-time delivery |
| Profitability: economics that are engineered, not hoped for | Margin, collections, cost per delivery |
Each pillar shows up twice: as a scorecard in Measure, and as the engines that produce it in Do. That pairing is what connects what you measure to what you do.
The memory underneath. What the machine reads to do its work.
Below the layers sit four structures. In most firms they barely exist. In an AI-native firm they're the foundation, because they turn tacit know-how into something a new hire, or a machine, can run.
- Value engines
- The defined, owned motions that produce each pillar: marketing and bookings, project governance and account management, revenue.
- Systems
- Connected enough to hold one shared source of truth, so the layers stop keeping separate versions of the same reality.
- Playbooks
- Living definitions of how each engine runs. The machine executes them and notices when the real work has drifted from the written version.
- Resource matrix
- A live map of who can do what, read by staffing and delivery decisions instead of guessed at.
Where Pulse fits. The first engine, built to the blueprint.
Pulse is the Leveraged Sales pillar made live. It reads your CRM and keeps pipeline, win rate and deal velocity current without anyone assembling them. Sellers get the Do layer: what to work, what they owe, what to send. Leaders get the Measure layer: what's real, what's at risk, where to step in.
It also changes the wiring. When everyone arrives at pipeline review already current, the meeting stops being about establishing what's true and starts being about what to do.
- Code computes the numbers. People make the calls.
- Every figure traces to a record.
- The method is written down, so it survives turnover.
The same pattern applies to every engine in the company: project governance and account management for delivery, revenue and collections for profitability. Each one starts the same way, by capturing how the work really runs today before designing what it should become.
Automation is only as good as the structure it runs on. We build the structure first.
How we decide what to build. Five filters, out loud.
- 01Is this getting us closer to our goals?
- 02Does this align with our values?
- 03Is this good for our customers?
- 04Is this something we can actually do?
- 05Does this make the company less dependent on any one person?
The fifth is the one most firms leave out. It's the test of whether you're building a company or a collection of people.
Design it.Measure it.Then automate it.
Start with one engine and one number. We'll show you the structure it runs on.