CarpaTech
Client Portal
AI, automation and data
AI Implementation
Custom Copilots and assistants. Document and email automation. Secure Azure OpenAI. Guardrails and evals.
Use-case workshops with the people who do the work. An honest look at your data and your processes. Risk and return scored per use case, and a 90-day roadmap at the end.
The model is rarely the hard part.
The hard part is upstream. The process is undocumented, the data sits in four systems with four spellings of the same customer, and nobody has agreed who signs off an answer the machine gets wrong. An assessment is finding that out on paper rather than finding it out in a build, and it ends with a ranked list rather than an opinion.
What the assessment produces.
A fixed scope with a document at the end. You can act on it with us, or with somebody else, or not at all.
- Use-case workshops with the people who do the work, not only the managers
- A data maturity audit: where it lives, who owns it, whether it agrees
- A process review of the candidates, written down as they actually run
- Risk scoring per use case, covering accuracy, privacy and regulation
- Return scoring, in hours or cost, with the assumptions stated
- A ranked shortlist, including the ones we recommend not doing
- A 90-day roadmap with the first build scoped and sized
- A written report you own, whoever does the work next
How the assessment runs.
We listen
Workshops with the teams whose work would change. The best candidates are usually named by the person doing the task.
We check the ground
Whether the data and the process can actually carry what was proposed. This is where most ideas get re-ranked.
We hand you a plan
Ranked use cases, scored, with a 90-day roadmap and the first build sized. Including what to leave alone.
A fixed-scope project.
The assessment is quoted as a project with a fixed scope and a document at the end, on any plan or without one. It is deliberately separable from the build work that might follow, so the recommendation is not written by somebody who has already been paid to implement it.
Projects, migrations and build work are quoted separately and covered by the same SLA.
The rest of this group.
The other three in this group. None of them works well until the foundations underneath can carry it, which is the order we build in.
Being asked what your AI plan is?
The honest answer starts with what your data and processes can carry. A short call and we will tell you whether an assessment is worth it. No audit fee, no obligation.