Innovation Accelerator Hub
All playbooks
Operating model stages

Readiness

Take the validation evidence pack through the Enterprise AI Readiness Review — security, scale, ops, cost, governance — and produce a readiness report that decides whether the bet is ready to hand off.

Published
Owner · Director, InnovationUpdated · 2026-07-21

Purpose

Readiness turns a validated bet into something the enterprise can actually take on. Validation answered should we invest. Readiness answers can we run this safely, at scale, without surprising legal, security, or ops on day one. It's a structured review, not a rubber stamp — and it can send a bet back for more work.

Entry
Validated bet with an invest recommendation from the validation debrief.
Artifact
Readiness report — one page per dimension, with cleared / gap / blocker status.
Gate
Cleared for handoff, or gaps named with owners and a target date.

The six dimensions

Run the prototype and the evidence pack through the Enterprise AI Readiness Review. Every bet is assessed against the same six lenses. Not every dimension has to be green — but every red has to be named with a plan.

Data & access
Security
What data does it touch, where does it live, who can see it, and how does it authenticate? Any PII, PHI, or regulated data changes the answer materially.
Load & performance
Scale
What does 10× the prototype's traffic look like? Where does it break first — model calls, database, third-party APIs, cost per request?
Run the thing
Operations
Who's on call, what's the monitoring story, how do we know when it's broken, what's the rollback? A prototype with no ops plan is a future incident.
Unit economics
Cost
Cost per user, per session, per inference. Extrapolate to real volume. Flag anything where the business case only works at prototype-scale pricing.
Policy & risk
Governance & compliance
AI use policy, model card, data-handling agreements, regulatory review (SOC 2, HIPAA, GDPR — whichever applies). Loop legal in early, not at handoff.
Who owns it
Organizational readiness
Is there a receiving team with capacity? Do they have the skills? What training or partner support do they need in the first 90 days?

Activities

  • Assemble the evidence pack. Everything the review needs in one place — prototype URL, validation summary, decision recommendation, architecture sketch, data inventory, cost estimate.
  • Run the readiness review. Score each of the six dimensions: cleared, gap (fixable pre-handoff), or blocker (must be resolved before the bet moves forward).
  • Loop in the right partners early. Security, legal, and infra don't like being surprised at handoff. A 30-minute conversation now saves a two-month escalation later.
  • Write the readiness report. One page per dimension. Status, the evidence behind the status, gaps, owners, target dates.
  • Decide: cleared, conditional, or back. Cleared → advance to Handoff. Conditional → advance with named gaps in the handoff package. Back → the bet returns to prototype or validation with specific work items.

Common failure modes

Treating readiness as a checkbox
A readiness review that always comes back green is a review that isn't reviewing. Every bet has real gaps — the report's job is to name them, not hide them. A conditional pass with three named gaps is more valuable than a clean pass that surprises the owner in month two.
Discovering security or legal late
If security and legal see the bet for the first time at Readiness, you've already lost weeks. Bring them in during Validation — even a short heads-up conversation dramatically reduces the odds of a blocker showing up here.
Confusing readiness with delivery
Readiness is not building the production version. It's naming what would need to be true for the receiving team to build it safely. If Readiness starts producing shippable code, the Accelerator has become the owner — that's Handoff's failure mode, one step early.

Supporting tools