Learning center · AI engineering acceleration
One ticket.
A better way to work.
Bring a piece of work that keeps getting stuck. Sit down with your team. Learn to specify it, build it, check it, and share what helped.
On the desk today
“Invite a colleague to a workspace.”
A small ticket with real decisions: who can invite, what happens on a retry, and how to check the workspace boundary.
Follow the worked examples →Before generation
Agree these first.
- Only a workspace administrator can invite.
- A retry must not create another invitation.
- Test access from a different workspace.
One ticket. Four working documents.
Public chapters. Practice with a coach.
Read and use every chapter freely. A coached cohort adds working sessions, feedback on your artifacts, help with obstacles, and phase reviews. Start with a pair of engineers, product and quality partners, and a sponsor who protects their time.
The suggested sequence is preparation followed by four working weeks. Agree the schedule with your coach; demonstrate readiness before progressing. Reading a page does not complete a phase.
The program at a glance
One shared practice. Five phases.
- Phase 0 · Suggested week 0
Prepare
Pod, backlog, access, baselines, tools, and structure.
Explore phase - Phase 1 · Suggested week 1
Execute
Agree the design together before generating a change.
Explore phase - Phase 2 · Suggested week 2
Guardrails
Keep changes reviewable and run the quality gates.
Explore phase - Phase 3 · Suggested week 3
Trust
Use AI to understand, explain, and debug the code.
Explore phase - Phase 4 · Suggested week 4
Scale
Transfer the playbook with coaches and rollout waves.
Explore phase
How performance builds, step by step
Remove waiting, reduce rework, make review repeatable, build shared understanding, then transfer the practice. Track accepted delivery and quality against your starting baseline at every step.
- Phase 0 · Prepare
Remove the friction before the work begins.
- Phase 1 · Execute
Spend less time on avoidable rework.
- Phase 2 · Guardrails
Turn faster generation into accepted delivery.
- Phase 3 · Trust
Investigate faster. Keep ownership.
- Phase 4 · Scale
Grow the gain across more teams.
Bring the measured results into the organization-wide rollout model: 5 trained by Month 1, 30 trained by Month 2, 50 trained by Month 3, with 30% gain per trained engineer.
Your path through the program
Welcome + five phases + outcomes- Welcome
Get started: Write your starting note
Name one friction, one ticket, and the people who can help.
Open chapter → - Phase 0
Prepare: Create the conditions to begin
Agree the scope, verify access, and establish your baseline.
Open chapter → - Phase 1
Execute: Turn intent into accepted work
Specify, generate, inspect, and validate a bounded change.
Open chapter → - Phase 2
Guardrails: Put the toolkit to work
Use shared skills, real checks, and a Proof scorecard.
Open chapter → - Phase 3
Trust: Investigate unfamiliar behavior
Trace source evidence, test a hypothesis, and teach it back.
Open chapter → - Phase 4
Scale: Transfer the practice
Coach a receiving pod and evaluate the next rollout wave.
Open chapter → - Wrap-up
Outcomes: Plan the next wave
Bring the practice together and explore a 50-engineer value example.
Open chapter →
Working on a specific activity? Open the Practice guides. For questions about participation and progression, read the FAQ.
A closer look
The work changes. So does the team.
Engineers spend more time agreeing intent, checking behavior, and helping another pod repeat the practice. Product, QA, and coaching make that possible.
Read about the new team structure →Work with our team through the program. Read our approach to kickoff and reviews.
Take the first small step
Name a real engineering friction and the people who could investigate it. The welcome chapter turns that into a starting note for your team or coached kickoff.
Write your starting note →