Advice from people who have actually shipped the work.
Most AI advice in this industry comes from people who have never had to deliver a finished master. Trained to Imagine consults for brands, agencies, and studios working out where generative tools genuinely fit — readiness audits, model and vendor selection, hands-on team training, and provenance and clearance policy. This is the advisory side of the business: we tell you what to do and teach your team to do it. When you want it built and run, that is the hybrid pipeline engagement instead.
AI readiness audit
An honest assessment of where generative tools would actually save money or unlock work in your organisation — and, just as usefully, where they would not. Delivered as a written recommendation with costs and risks, not a maturity matrix.
Model and vendor selection
Which models for which task, at which budget tier, with which licensing and provenance position. Our founder built tiered model strategies across a wide range of image and video models while running production at scale.
Team training and workshops
Hands-on sessions with your creative team on real briefs of yours, not demo prompts. People leave able to do the work, which is the only measure that matters.
Provenance and clearance advisory
Policy for legal and brand teams on training data, likeness, rights, and documentation. Recent rulings have moved this from reputational concern to licensing obligation, particularly in Germany and the EU.
- 01
Scope the question
What decision are you actually trying to make? Most engagements start vaguer than they need to be.
- 02
Look at the real work
Your briefs, your budgets, your approval chain, your delivery specs. Generic advice comes from generic inputs.
- 03
Recommend and prove
A written recommendation, and where it is contested, a small piece of real output that settles the argument.
- 04
Enable the team
Training and documentation so the recommendation survives our leaving. If it needs us to work, it has failed.
What does AI consulting for a creative company actually involve?
In practice, four things: working out where generative tools fit your existing process, choosing which models and vendors to use for which task, training your team to operate them, and setting policy on rights and provenance. What it should not involve is a technology roadmap detached from the work you actually ship. If the advice cannot be tested against a real brief on a real deadline, it is not worth paying for.
How is this different from your pipeline engagement?
Consulting is advisory: we assess, recommend, and train, and your team executes. The pipeline engagement is hands-on construction: we design the workflow, build the tooling, run a pilot project through it, and hand over documented standard operating procedure. Many clients do the consulting first and the build second. Some only ever need the first.
Do we need a consultant if we already use AI tools?
Often the opposite problem applies: teams already using these tools have accumulated undocumented practice, inconsistent model choices, and no record of what produced what. That is fine until a client, a broadcaster, or a legal team asks. An audit at that stage is usually about consolidating and documenting what already works rather than starting over.
Can you advise on AI rights and clearance for the European market?
Yes, and it is the fastest-moving part of this work. The Munich court's ruling against Suno established that systematic commercial use of copyrighted training material requires licensing, which shifts the question from what a model can produce to what it was built on. We advise on documentation practice and model selection against that constraint. We are not lawyers and do not replace counsel — we make sure your production process can answer the questions your counsel will ask.