Work stream
AI delivery model
The single end-to-end way we deliver work with AI, captured as one clear, shared model that we all use internally and with our clients.
Finalise wording & brief Katya
- 4 stages (Set up, Explore, Deliver, Optimise/Support)
- Human, AI-assisted & agentic
Katya to design visual of model (in Ah Flow styles)
Review & agree v1
The agentic offering
Agentic capabilities that transform delivery for our clients and amplify our team, letting our people direct agents to do far more, far faster at a higher standard.
Capability workshops, by team
- Design - Fergal
- Dev / QA - Kevin
- PM - Nick
- Digital Performance - TBC
- Analytics & others - TBC
Consolidate into one agentic offering
Review & agree v1
Team / pod structure & costing
How a team is organised to deliver, and what a pod costs. Pricing and validating costs vs salaries sits with John & Co.
Confirm the lean pod (Orchestrator / Creative / Build)
Map disciplines into the three roles at different scales
✓Pod cost model
Define how existing teams relate to the pod
Hand structure + costing to commercials for pricingthen John & Co
AI in live engagements
The live proof across real client work: FITRADE, FICS, Signode, An Post and others. Anchors dates; does not gate the model.
FITRADE Stage 2 Component Dev (in window) + capture evidence
Identify next engagements to bring on (FICS, Signode, An Post)
FITRADE Stage 3 begins · economics + quality evidence
Internal AI rollout
Getting the teams using it: the AI Journey, onboarding, champions, upskilling and internal comms.
Publish the Ah Flow space + AI Journey to the team
Name champions; agree the upskilling path per role
Run onboarding + Share & Learn sessions
Track adoption across teams
Learnings & feedback
Capture what works and breaks across every team using the model, and feed it back to refine it.
Set up the cross-team feedback loop + retro cadence
Capture learnings from live work + internal use
Feed refinements back into the model