Approach · Agents · 9 min read
Why 80% of enterprise AI pilots never reach production — and the seven gates that fix it.
The graveyard isn't full of bad models. It's full of good ones that never got an owner, a control, or a financial baseline. Here is what we apply to every engagement — including our own.
Most enterprise AI doesn't fail at the model. It fails before the model gets a chance to operate in production. The pattern is so consistent that, when a prospect walks in with a stalled pilot, we already know which of the seven gates was missed.
The graveyard is full of good demos.
A demo proves a model can answer a question. Production requires the model to act inside a business process — with approvals, exceptions, audit trails, and someone who can explain to a regulator why it did what it did. The work between the two is where AI programmes stall, and where most vendors stop showing up.
The seven gates.
We apply the same seven production gates to every agent we build. We also apply them to our own internal agents. If a use case cannot pass all seven, it does not ship.
- 01
A named business owner
Someone in the business accepts accountability for the outcome and operating risk — not the vendor.
- 02
Documented process & approvals
The workflow, human approvals and exception paths are written down, reviewed and versioned.
- 03
Approved data & access
Sources, access rights, retention and permitted use are signed off before any model touches them.
- 04
Defined & tested targets
Accuracy, safety, reliability, latency and cost targets are specified and measured against evaluation sets.
- 05
Least privilege by default
Agent permissions, scopes and transaction limits follow least privilege — and are reviewed regularly.
- 06
Operational controls
Monitoring, logging, incident response, rollback and shutdown controls are in place before go-live.
- 07✓Applied to every engagement — including our own.
Financial baseline
A baseline and benefits-measurement plan are agreed before launch — so success can be evidenced.
Why these seven.
Gates one, two and seven are about accountability: who owns the outcome, how is the work described, and how do we know it paid back. Gates three, four and five are about evidence: data, performance, and authority. Gate six is about operating reality — what happens when something goes wrong at 02:00.
The single biggest predictor.
In our experience the single biggest predictor of an agent reaching production is whether gate one — a named business owner — exists at the start. If no one in the business owns the outcome, the agent will be everyone's pilot and no one's product.
Want a one-page view of how this looks on an engagement?
Book an executive briefing — we'll walk you through the gates applied to a real production agent, and what would need to be true for your pilot to ship.