Stage 3 · Forge

Commissioned. Momentum-locked.

Pick one or more tracks. Forge runs to completion before another department's Ignite — that's how the flywheel stays honest.

8–12 weeks

Deep Learning Track

Adaptive — adjusts to maturity per role.

Who's involved
Whole department, role-tailored
Deliverables
  • Role-based AI curricula
  • Hands-on labs
  • Certification milestones
  • Manager toolkit
The rule. Forge tracks must be commissioned and underway before another department's Ignite begins. It protects quality, and it protects The Commons from drift.
Who builds it

Your use cases. Your choice of builder.

Once a use case is discovered, prioritised and business-cased, delivery can go three ways. The pattern, guardrails and architecture stay the same — the hands can change.

Option 01
Your own delivery team

The department builds it in-house using the Commons: reference patterns, guardrails, evaluation harnesses and reusable components. We coach, review and assure — you own the keyboard.

Option 02
Your chosen partner

Keep your incumbent SI, local supplier or framework vendor. Ascend·Gov stays as the architecture and assurance layer so what they ship still fits the estate and passes the same gates.

Option 03
Built with Avaiden

Ascend·Gov builds it with Avaiden, an AI-native product, engineering and delivery studio: agentic AI embedded across the whole lifecycle — discovery, design, code, test, release and run — with senior pods and MVPs measured in weeks, not quarters.

avaiden.com →

No lock-in, either way. The government owns the IP, the data and the deployment. Every build lands in The Commons so the next department starts from a working pattern rather than a blank page.

Adoption is built in

Every Forge track ships with an adoption matrix.

Awareness · Capability · Confidence · Habit — measured per role, not just at programme level.

See how it feeds The Commons