bmax

Every person at Pursuit has access to premium AI tooling. Engineers, sales reps, customer success — everyone. Add in voice-to-text, MCP integrations, analytics, project management, and we're spending meaningfully on AI per employee per month.

My policy: unlimited budget for any AI tool that improves productivity. If it saves time, we buy it. No approval process, no justification needed. The ROI is so obvious it doesn't need a spreadsheet.

What changed

This isn't just an engineering story. AI changed every function at the company.

Engineering doesn't look like it did a year ago. Our engineers spend more time testing, reviewing, and making product decisions than writing code. One engineer on our team put it well: "If I have to write a curly brace now, I feel pain." Every keystroke an AI can handle is a keystroke a human shouldn't be doing.

Sales runs AI-powered targeting that dramatically outperforms manual prospecting. Every rep asks AI-discovery questions on calls. We use AI to qualify prospects before a trial even launches — so the team spends time on deals that are actually going to close, not ones that look good on paper.

Product uses AI to digest usage data and customer feedback at scale — surfacing what actually matters instead of what's loudest. We found blind spots where the team was spending the majority of its energy on things that weren't driving churn. AI didn't just speed up prioritization. It changed what we prioritized.

Customer success uses Claude projects for onboarding — deep account research, company profiles, configuration review. Meeting notes get piped through AI for structured analysis so nothing falls through the cracks. The team moves faster with better context on every account.

AI on the org chart

The bigger shift isn't how people use AI tools. It's that AI fills roles that used to require people.

We have agents that run nightly — reading company profiles against millions of government documents and surfacing procurement signals. We have multi-step AI verification pipelines for contact accuracy. We have agents that generate outbound sequences, process public records requests end-to-end, monitor news across tens of thousands of entities, and classify documents at scale.

These aren't features. They're jobs. They show up on our org chart the same way a data analyst or research associate would. The difference is they run at 3am and process a million documents before anyone's had coffee.

Customer-facing stays human

Here's the line we draw: automate everything the customer doesn't see. Keep humans everywhere they do.

The backend is fully automated — crawlers, document parsing, classification, signal scoring. But when a customer gets on a call, that's a person. When someone's onboarding, that's a person with AI-powered context, not an AI with a script. When a rep sends a follow-up, they're editing and personalizing what AI drafted — not copy-pasting.

AI handles the grunt work. Humans do the things that actually matter in enterprise sales: building relationships, understanding nuance, making judgment calls that require context no model has.

The new org chart

The future org chart has two columns: people and agents. Both have costs. Both have output metrics. Both need to be managed.

Companies that add AI to their team will outcompete companies that replace their team with AI. The companies that figure out this balance first will build the most durable businesses of this era.