We reject any project we cannot explain without saying AI
At Omniance, we reject any project we can't explain without saying "AI".If it needs buzzwords, it's weak.
Last year I watched shiny "AI first" pilots die in silence.Nice demos, no impact, no link to profit.Everyone praised the tech stack, nobody could state the business case in plain words.
I didn't want to do the same mistake.
We hard coded a rule into our first call with a client:no AI language in the problem statement.
We start every engagement with three steps:
- Describe the pain in simple terms a CFO would accept
- Map the real process, from trigger to outcome, with names, systems, handoffs, data quality
- Link each step to P&L logic, euro by euro
Only after that work do we talk about models, architecture.
At times the result is clear.A rule based fix solves the issue.Or a small workflow change removes half the waste.No AI required.
For me this is a leadership principle, not a sales trick.
We must think like capital allocators.Each euro into AI must work harder than the next best use of that euro.AI becomes a consequence of a good case, never the start.
Clients stay out of innovation theatre.Our team stays out of tool shopping.
The race for more AI is loud.The quiet race for clear problems decides who wins.
Do not buy the illusion of "AI first".Force clarity on the problem first.