AI Readiness Assessment
Two weeks to separate the AI use cases worth building from the ones that only demo well, scored against your actual data and workflows.
What you get
- A use-case shortlist, scoredCandidate workflows ranked by value and feasibility against your real data and systems, not against a vendor demo.
- Data readiness findingsWhether the documents and records the use cases depend on are accessible, clean enough, and safe to expose to a model.
- ROI estimates per use caseHours saved and error costs avoided, stated with their assumptions so finance can argue with the inputs, not the math.
- Governance gapsWhat an AI use policy, data classification, and access boundaries need to cover before the first deployment.
- A build sequenceWhat to pilot first, what it should cost, and what to deliberately not build yet.
What we need from you
- Interviews with two or three process owners (about three hours total).
- A tour of the systems and data sources involved (read-only or over a screen share).
- Any AI experiments already tried, including the failed ones. Especially the failed ones.
Timeline
Interviews, workflow inventory, and data source review.
Scoring, ROI modeling, governance gap review, and the written report with walkthrough.
Who it's for, and who it isn't
- The board keeps asking about AI and you want an answer grounded in your operation.
- Teams are already pasting company data into public chatbots and you need a sanctioned path.
- You have budget for one pilot and cannot afford to pick the wrong one.
- You have already chosen the use case and want it built. Skip the assessment; talk to us about the build.
- You want a slide deck that says AI will transform everything. The report says what the numbers say.
Sample deliverable: table of contents
- 1 · Executive summary
- 2 · Workflow and use-case inventory
- 3 · Scoring: value vs feasibility
- 4 · Data readiness findings
- 5 · ROI estimates and assumptions
- 6 · Governance gaps
- 7 · Recommended build sequence
Common questions
Do we need to clean up our data first?
No; finding out what your data can actually support is part of the assessment. If the honest answer is that the data platform comes first, the report says so and sequences it.
Which AI vendors do you favor?
None by obligation. We build on Anthropic Claude, OpenAI, and the cloud-hosted variants, chosen per use case on quality, cost, and data-handling terms, and we hold no reseller arrangements that would tilt the answer.
Do you build a pilot as part of the assessment?
No; the assessment exists to make the pilot a safe bet rather than a gamble. The build is a separate engagement, and the report prices what the first one should cost.
What if the answer is that AI does not fit us yet?
Then that is the finding, along with what to fix first and what signal should trigger a second look. A report that always recommends building would not be worth buying.
Related
Connect with us about where AI actually fits.
Two weeks, a scored shortlist, and honest ROI math from a team that operates its own AI SaaS in production.