AIM Network Hub is the shared learning layer installed on aim-engine.com. Customer AIM Engine installations connect to it without handing over their OpenAI API keys.
The feedback loop
Customer campaign → structured evaluation → selected sharing scope → AIM Network Hub → SCAMPER improvement → shared Network Library → adoption by another campaign → new measured outcome.
What reaches the Hub
Network identity
Connected company/site identity, plan and last activity.
Aggregate activity
Input and output counters can be synchronised without sending private prompt text.
Shared knowledge
Completed research and company updates deliberately submitted to the community.
Campaign evaluations
Structured fields for objectives, inputs, activities, outputs, out-takes, outcomes, impact, stakeholders, confidence and SCAMPER analysis.
Agile learning states
Incoming → Classified → In progress → Measurement → Outcome received → Impact review. The Hub records the evaluation history so a campaign can evolve as real evidence arrives.
What comes back to customers
The Network Library can return relevant approved research, anonymised learnings and improvement candidates. Customers can save an item to their own Library, adopt it into a Project or Campaign, test it and later return the outcome.
Why this is different from generic AI memory
The network can build a genealogy from evidence to idea to adoption to outcome. A strategy is therefore not only “something AI suggested”; AIM can retain the context of which earlier learnings contributed to it and whether later campaigns produced evidence for or against it.
Private work remains a first-class state.
The network effect should come from deliberately shared learning, not from silently publishing each customer’s working data.