Many companies know AI could help. Few know where to start. The expensive path is a free “discovery workshop” that ends in enthusiasm and no artefacts — then a large SOW justified by vibes.
A proper AI Opportunity Assessment is a paid standalone product. You pay for a concrete automation plan with an economic case, not for the privilege of being sold later.
This article is for CEOs, COOs, CTOs, and Heads of Operations who want to implement AI in operations or existing systems — the ops track of AI Transformation (engineering SDLC transformation is a parallel track).
Why paid beats free pre-sale
Free discovery incentivises the vendor to sell the next phase, not to kill bad ideas. A paid assessment aligns incentives: the deliverable is the plan. You can take it to another partner, build in-house, or buy the next engagement with eyes open.
D-Factor treats assessment as AIT·ASSESS — entry to workflow automation and later scale — not a complimentary workshop.
What you should receive
For a 1–3 week engagement, expect artefacts like these:
| Deliverable | Why it matters |
|---|---|
| Analysis of 3–5 business processes | Enough breadth without boiling the ocean |
| Automation potential per process | Separates hype from leverage |
| Prioritised AI use-case list | Forces ranking, not a wishlist |
| Preliminary architecture sketch | Systems, data, HITL boundaries |
| Cost, risk, and ROI view | Finance can challenge assumptions |
| Implementation roadmap | Sequence: pilot → workflow → agents / scale |
Outcome: a written AI automation plan you can defend in a leadership meeting.
What “good” looks like in the analysis
- Owner named for each process (ops, not only IT)
- Volume and cost of today’s manual path (even ranges)
- Risk class — where human-in-the-loop is mandatory
- Data readiness — access, quality, residency
- Integration reality — CRM / ERP / APIs, not chat-only demos
If the report only lists “chatbot for support” with no volumes or systems map, it is marketing.
How assessment connects to delivery
Typical path after a useful assessment:
- One workflow (
AIT·FLOW) — automate a single process end-to-end - Agents & integrations when tool-calling becomes the bottleneck
- Enterprise multi-process only after pilots prove ROI
- Managed operations after go-live — who owns quality and cost of agents → Managed Application Services / AI agent operations
Engineering tool rollout (coding assistants, SDLC policy) is a different problem — see the AI Engineering Transformation page. Do not conflate ops automation with eng enablement in one fuzzy SOW.
Buyer checklist before you sign
- [ ] Assessment is priced and scoped as its own product
- [ ] Deliverables listed in the SOW (not “workshop + recommendations”)
- [ ] 3–5 processes agreed upfront
- [ ] ROI / cost section required, even if ranges
- [ ] You keep the artefacts regardless of who builds next
Conclusion
Paid discovery is how serious AI programmes start. Free slides create free confusion.
If you want a prioritised plan with a business case — not a demo — start from an AI Opportunity Assessment under D-Factor’s AI Transformation work, then decide whether the next step is a single workflow, engineering enablement, or a dedicated delivery unit.