Built This Week/Builds

Builds

Monte Carlo Sales Forecast Dashboard

Sam turned Attention's open-source, one-call-at-a-time coaching scores into a 10,000-trial Monte Carlo forecast of the whole pipeline, and Attention's CEO critiqued it live.

For Episode 45 (June 5, 2026), Sam spent a day building on top of the sales coaching stack that Attention had just open-sourced. The result is a pipeline dashboard that runs a Monte Carlo forecast from conversation scores and puts it head to head with what the team committed in the CRM. Then Attention's co-founder and CEO, Anis Bennaceur, picked it apart on air. Everything runs on fake data.

Why we built it

Attention's open-source release scores sales conversations one at a time. Sam wanted to take the next step: roll those per-call scores up into a forecast for the whole pipeline. What a CRO wants, he said, is as small a gap as possible between the forecast and the CRM commit. He opened with a warning that it "may be completely off base," since he had little familiarity with Attention's tool.

The stack

  • Attention's open-source sales coaching stack: the conversation scoring and coaching reports the dashboard is built on.
  • A Monte Carlo simulation layer Sam added on top: 10,000 trials over the pipeline.

Sam didn't say on the show which coding tool he used for this one.

How we built it, step by step

  1. Start from Attention's open-source stack. It already knows the rep, the buyer and their roles, links to the CRM, and produces coaching reports and scores for each call.
  2. Build the pipeline view. The headline compares the AI forecast, $1.17M, with the CRM commit, $1.44M.
  3. Add deal drill-downs that explain risk. One deal, Helios, looked strong until a new board-level vendor risk review gate appeared two weeks earlier.
  4. Add a live call coaching view. It scores the rep while the call happens. In the demo, "Jordan" as the rep scored 67 out of 100.
  5. Add scenarios. A macro tank, and the Helios deal falling through, each show how the gap to the CRM commit changes.
  6. Run the Monte Carlo replay. 10,000 trials based on the conversations to estimate how close the pipeline is to the forecast.

How it turned out

Anis liked it more than Sam expected:

And so I'm pretty impressed with what you did here in such short amount of time.

— Anis Bennaceur, Episode 45

He said Attention had never done Monte Carlo forecasting and had "never even thought about it," though he had used something similar to decide go or no-go on automated outbound messages. He also liked that the forecast landed relatively close to the CRM commit, which matches what most companies see, and that deal risk came from the coaching frameworks and the actual conversations.

His critiques were specific. Use real sales terms for the bands: commit, best case, at risk. And the feedback text sounded "very AI-ified." He mentioned that he had also released a humanizer, and that telling it to write like Paul Graham made it read like a real person.

What we'd do differently

The biggest lesson came from Anis at 06:53: forecasting is only one part of what leaders care about. The other is what to do about it.

  • Go from forecasting to actions. If a deal is at risk because you haven't reached the real decision maker, the tool should offer a button that drafts the multithreading message. The rep reviews and sends it. That improves the forecast instead of just reporting it.
  • Speak the CRO's language. Label deals the way sales teams already do.
  • Humanize generated coaching text so reps actually read it.

For more on how AI is changing sales teams, see AI in Sales. Another scenario-style build of ours is the portfolio macro scenario simulator.

FAQ

How do you build a Monte Carlo sales forecast with AI?

Sam started from Attention's open-source sales coaching stack, which scores each conversation, then used those scores to run a 10,000-trial Monte Carlo forecast of the whole pipeline and compared it with the CRM commit. He added deal drill-downs, a live call scorecard and macro scenarios.

How does an AI pipeline forecast compare to the CRM commit?

On fake demo data, Sam's AI forecast was $1.17M against a $1.44M CRM commit. Anis Bennaceur said it landed relatively close to the commit, which is what most companies see.

What is Attention's open-source sales coaching stack?

Attention CEO Anis Bennaceur open-sourced a dialed-down version of what Attention does. It scores sales conversations one at a time. Attention had never done Monte Carlo forecasting before Sam's build.

Is forecasting enough for sales leaders?

Anis said no. Forecasting is half of it; the other half is proactive actions that save deals, such as drafting a multithreading message to an unreached decision maker for the rep to review and send.