Training · AutomatorMoveemergingverified

Build Clinic

also known as AI clinic, automation drop-in, AI office hours

A recurring session where participants bring one real task and leave with a working automation. The goal is a concrete time-saving tool produced in the session, not a demo.

How the learner advances

flowchart TD A[Participant nominates recurring task 24h ahead] --> B[Facilitator scopes task to single session] B --> C[Participant builds automation with AI/no-code tools] C --> D{Live test on real data} D -- Fails --> C D -- Passes --> E[Demo to group + brief retro] E --> F[Post prompt + blueprint to shared library] F --> G[Automation runs in production next week]

Intent. Turn one real recurring task per participant into a working automation inside a single facilitated session.

When to apply. Use when a team has adopted AI tools but individual workflows remain manual, and short peer sessions are more realistic than a formal course. Especially effective when different team members have different tasks — the bring-your-own-problem format keeps every session immediately relevant.

Threshold — earns the next step. The participant can operate the automation they built without facilitation assistance and can explain what trigger, action, and edge case it handles.

Masterpiece — the artifact that proves it. A deployed automation that eliminates a specific recurring manual task, has passed a live test on real data during the clinic, and is documented in the team's shared library for others to adapt.

Example scenario

A marketing coordinator at a 60-person SaaS company attends a Tuesday Build Clinic. She nominated her task the day before: every Monday morning she manually copies campaign performance numbers from three spreadsheets into a summary email for her manager. During the 90-minute session she builds a Make scenario that pulls from Google Sheets, formats a summary, and sends it via Gmail at 8 AM every Monday. The facilitator helps her fix a date-formatting issue without touching her keyboard. At the live test, the scenario fires correctly against last Monday's real data. She adds the blueprint to the team library and tags it 'reporting'. Her manager notices the automation the following Monday and asks her to extend it for two other reports. The clinic produced one capable builder, one running tool, and one reusable recipe — all in under two hours.

Facets

Container — build-clinic
Mode — byo-problemhands-on-buildmentored-unblock
Reach — team
Persona — analyst-opsnon-technicalmanager-leader
Craft (AI Fluency) — delegationdescriptiondiligence
Learner — human
Trainer — human
Guardrail — responsible-use

Inputs

Nominated task — A specific recurring task submitted by each participant at least 24 hours before the session, scoped to something buildable in under two hours.
Facilitated build environment — A shared workspace with access to no-code tools (Zapier, Make, n8n) or AI assistants (Claude, ChatGPT), plus a facilitator who can unblock participants live.
Shared prompt or recipe library — An existing or nascent library where finished automations are deposited so the team benefits beyond the session.

Outputs

A more capable learner — A participant who has personally built and tested one automation and can repeat the build process independently for their next task.
Working automation (Masterpiece) — A live, tested tool that handles a specific recurring task, verified in the session and saving measurable time within one week.
Library contribution — A documented prompt or automation blueprint added to the team's shared library within 24 hours of the clinic.

Steps (3)

Pre-session task nomination
Participants submit one recurring task at least 24 hours before the clinic. The facilitator reviews submissions and helps anyone whose scope is too large to fit a single session, trimming the problem to a buildable slice.
Scoped build
Each participant builds their own automation using no-code or AI tools chosen to match the task. The facilitator floats, unblocking stuck participants without taking over — the learner drives. The session ends when the tool passes a live test on real data.
Live test and retro
Participants demo their automation to the group, noting what worked and what needs refinement. Retro is brief (5 minutes per participant); the goal is capturing one improvement action per tool. Finished prompts and blueprints are posted to the shared library before the session closes.

Principles

Every participant ships something that runs on real data before leaving.
The facilitator unblocks; the learner drives — taking over to finish for someone erases the learning.
A working tool, however small, beats a polished demo of something not yet running.
Library contributions compound team capability; skipping the deposit wastes the clinic's value.

Unlocks methodologies (2)

A learner who completes this pattern is equipped to execute these methodology families:

Prompt Engineering Iteration Management

Known uses (3)

Known failure modes (3)

[facilitator-takeover]
The anti-pattern of the facilitator finishing a participant's build when they get stuck. This erases the hands-on learning and leaves the participant unable to build independently next time.
[scope-creep-at-intake]
Accepting over-sized tasks without trimming at nomination stage means participants leave with half-built automations and no working tool — the opposite of the pattern's intent.
[no-library-deposit]
Finishing the session without depositing prompts and blueprints to the shared library. The anti-pattern converts a team-wide learning event into one-off individual wins that cannot be reused.

Related trainings (2)

Sources (3)

Provenance

Ecosystem: neutral
Added to catalog: 2026-05-27
Last updated: 2026-05-27
Verification status: verified