Before

Five people were using AI when we started. The executive assistants sorted the partners’ inboxes by hand, and presentations were pulled together from scratch each time.

What we built

1. A company AI brain that knows Cadence’s projects, templates and history.
2. Email sorting and first drafts in the assistants’ inboxes. It drafts, a person sends.
3. A shared library of skills anyone in the office can use.
4. Training for the whole team, and office hours every fortnight.

After

Seven weeks after kickoff the whole office was using it, and staff with no coding background had started building their own tools.

Before

The fund’s knowledge was spread across six Google Drives, Notion and its CRM. Partners were building AI tools by hand and spending their time fixing them.

What we built

1. Every night the brain reads the six drives, Notion and the CRM, and shows each person only what they are allowed to see.
2. Skills for the team’s regular jobs, like valuation prep, in one shared library.
3. Two of their own team manage the library, so new skills are checked before everyone uses them.

After

The whole team now uses AI every week. In the quarterly valuation cycle, the team leaned on a skill one of their own people built.

Before

Booking first-round interviews meant emailing back and forth late at night. For 10 candidates it could take a couple of hours.

What we built

1. An agent Tom talks to in Telegram, connected to his email and calendar.
2. He sends the CVs and the role as a voice note.
3. It emails every candidate, books each reply into his calendar and sends the confirmations.
4. It tells Tom who is booked and who has not replied.

After

Tom now books a round of interviews with one voice note. He puts the saving at five to 10 hours a week.

Before

Phil’s week went on repetitive weekly reporting and small ownership tasks, not on the floor with members.

What we built

1. An audit of where the owner’s week was going.
2. A roadmap of what to automate first.
3. The weekly reports, automated end to end.

After

The weekly reporting runs on its own. Admin costs are down 40%, and customer service scores are up by as much as 80%.

Before

A lean team was buried in the admin behind every booking and every client message.

What we built

1. One company AI brain for the whole team.
2. Automation across the admin behind every booking.

After

In Toby’s words, the team’s admin hours have quartered.