Customer Delivery Playbook

Purpose

A reusable shape for FDE-style customer work: discovery, prototype, implementation, deployment, and learning loop.

1. Discovery

Questions:

• What workflow is painful today?
• Who owns each step?
• What systems are involved?
• What data is required?
• What exceptions happen often?
• What does success look like?
• What would make the customer trust the system?

Outputs:

• Workflow map.
• Systems map.
• Data/source inventory.
• Risk list.
• First use case.

2. Prototype

Build the narrowest useful version:

• one input path
• one core workflow
• one integration or mock integration
• one AI-assisted step
• one human review step
• one output/dashboard

Outputs:

• working demo
• known limitations
• open customer questions

3. Implementation

Move from demo to usable system:

• real data shape
• auth/secrets
• error handling
• logs
• permissions
• setup docs
• customer feedback loop

Outputs:

• implementation checklist
• go-live checklist
• support/troubleshooting notes

4. Go-Live

Before go-live:

• confirm owners
• confirm data sources
• confirm rollback/disable path
• test failure cases
• test user review flow
• prepare support contact path

After go-live:

• watch logs
• collect user feedback
• track accepted/rejected AI suggestions
• identify repeatable patterns

5. Product Feedback

Capture:

• what was custom but should be reusable
• what confused the customer
• what data was missing
• what integration assumptions failed
• what docs or tooling would speed up the next deployment