AI Offerings
Fixed scope, fixed price, defined deliverable.
AI readiness assessment
An inventory of your data, candidate use cases and risk exposure, ending in a ranked roadmap with cost and payback for each item.
- Duration
- 2 to 3 weeks
- You get
- An inventory of the data that candidate use cases depend on
- Candidate use cases, ranked
- A summary of risk exposure for each
- A roadmap with cost and payback for each item
- You provide
- A sponsor who can make decisions
- Access to the people who own the data and systems
- Existing architecture and policy documents, if you have them
- Starting price
- Scoped on call
Production agent build
One workflow, one agent, shipped to production with governance and evaluation built in from the first release.
- Duration
- 6 to 8 weeks
- You get
- One agent running one workflow in production
- Policy checks and audit records from the first release
- An evaluation suite that runs on every change
- A runbook and a handover to your team
- You provide
- The workflow and the person who owns it
- Access to the systems the agent will touch
- Someone with authority to approve the policy
- Starting price
- Scoped on call
Model evaluation and red-teaming
A test suite for accuracy, drift and failure modes before launch, rerun on every model or prompt change.
- Duration
- 3 to 4 weeks
- You get
- A test suite covering accuracy, drift and failure modes
- Red-team findings, ranked by severity
- The suite wired to rerun on every model or prompt change
- You provide
- Access to the model or endpoint under test
- Examples of good and bad outputs
- Your definition of an acceptable error
- Starting price
- Scoped on call