Skip to article

What Is an AI Agent Build Session?

You build one, on your own account, live in the room.

By Jacqueline V. TwillieAugust 17, 20263 min readDefinition

An AI agent build session is a guided, live working hour where each participant builds a functioning AI agent on their own account, pointed at a real task, and leaves with it running.

Every person in the session has their own platform account open, follows the same five-move sequence, and ships a working AI agent before the hour ends. Nobody watches someone else build. The builds happen on Gumloop, which runs on a 14-day trial and then a paid plan.

What happens in the session?

Five moves, in order, same for everyone:

  1. 01

    Name the AI agent and give it a one-sentence job.

  2. 02

    Write the instructions in plain language, the way you would brief a new hire.

  3. 03

    Switch on only the abilities the job requires.

  4. 04

    Connect the AI agent to the tools it needs with scoped access.

  5. 05

    Run it on a real task and watch the output.

The most common first build is a Meeting Prep AI agent: it reads the calendar, pulls the attendee list and the last thread with those people, and hands the participant a brief before their next call. Participants test it on a real meeting and use the output that day.

How BNEDai defines this differently

Most AI training sessions teach concepts. Participants learn what AI agents are, see case studies, and leave with a slide deck. A build session inverts that order: the participant builds first and understands the concept through the build.

BNEDai's Agent Lab sessions have taught more than 300 people to build working AI agents across 29 live sessions, June 29 to September 30, 2026 (source: AI Agent Build Impact Report No. 1, BNEDai, updated October 2026). Of surveyed participants, 90.5% reported confidence to build the next AI agent on their own (21 respondents, Google Forms survey through August 14, 2026, same source).

Operating rules are part of the session itself. "Draft, never send" is an instruction written into the AI agent's brief during the build. Scoped access is a setting toggled during move four.

For the full five-move build sequence with detailed guidance on each step, see how to build your first AI agent.

Frequently asked questions

What is an AI agent build session?

An AI agent build session is a live, guided working hour where each participant builds a functioning AI agent on their own account, pointed at a real task, and leaves with it running.

Do I need to know how to code?

No. The entire build happens in plain language on a no-code platform. No programming is required, and the instructions are written the way you would brief a new hire.

What do I need before the session?

An account on Gumloop (14-day trial, then paid), the login for whatever tool your AI agent will connect to (usually a calendar or inbox), and one real task coming up this week that the AI agent can work on.

What is the most common first AI agent people build?

A Meeting Prep AI agent that reads the calendar, pulls the attendee list and the last thread with those people, and produces a brief before the call.

Can I attend more than one session?

Yes. Many participants return to build additional AI agents. The five-move sequence is the same each time; the job changes.

The next step

Build something that actually runs your workflow.

A focused, free 60-minute live session with Jacqueline. You build alongside her, on your own real task, and leave with an AI agent that is already running.

More on Vocabulary for References