Your team already bought the tools and the AI. Rulebook makes them do the work.

The tools sit idle

Revenue teams run on great software: a CRM, a call recorder, an enablement platform, a dozen more. Most of it sits idle. In Pendo's product data, 80% of features are rarely or never used, and just 12% of features get 80% of the clicks^[1]. Zylo's 2026 index finds that 46% of SaaS licenses go unused in a typical month^[2]. Companies pay for the whole platform and use a small part of it.

The tools are not the problem. The last mile is. A leader decides something in a meeting: follow up within a day, no discount above 15% without approval, update the forecast before Friday. Then someone has to turn that sentence into clicks across five systems, every week, without being reminded. Usually nobody does. The decision fades before the next meeting.

The AI sits idle too

Now the same thing happens with AI. Boards and markets push every company to become AI native. Leaders buy seats for AI assistants like Claude and hand them out. But nobody has time away from the day job to learn the technical side and keep up. In a 2025 BCG survey of more than 10,600 workers, only 36% said they feel adequately trained to use AI^[3].

So the most capable models in the world draft emails. The real work still waits for a person: the deal update, the follow-up, the forecast, the order check.

Why I build Rulebook

I have spent more than seven years on that last mile. I was the fifth GTM hire at Postman and helped build its enterprise motion from Series A to Series D. I joined DevRev before revenue and ran revenue operations and enablement through its first $10M in ARR. I led revenue enablement technology at ABBYY.

As an enablement leader and then a GTM engineer, I saw the same pattern at every company: smart decisions, said out loud, then lost between the meeting and the tools. Better software did not fix it. More AI seats will not fix it either. The work has to get done for people, not handed to them as one more thing to learn.

What Rulebook does

Rulebook understands the work. It reads your meetings, Slack and email, and finds each thing that someone must do. Each item gets an owner before Rulebook acts on it. It turns the work into a plain-English plan. By default, the owner approves a change to a record before Rulebook makes it. Nobody writes a prompt.

Then Rulebook does the work in the tools you already pay for. Today that is HubSpot, Salesforce, Slack, Gmail, and Google Drive, Docs and Slides. It reads each change back from the tool to prove it, and it keeps a receipt of each step. Every fact it writes has a source.

In development Next, Rulebook will pick the right connection for each job: an API, or an MCP server (a standard way for AI to use an app). Where a tool has no API, it will use a browser like a person does, one checked step at a time. We show these parts as live only after they run on real accounts.

Say it once. See it done.

Abhishek Humney, Founder and CEO

1. [1] Pendo, The 2019 Feature Adoption Report
2. [2] Zylo, SaaS Management Index 2026
3. [3] BCG, AI at Work 2025: Momentum Builds, But Gaps Remain, June 26, 2025