Frequently asked questions

The pre-assessment starter is free and can begin with a scan of public evidence. In the current founding offer, the Validation workshop is also free and deliberately bounded. For direct buyers, paid work starts with a single-company Opportunity sprint from $495, then a pilot handoff from $25,000 when engineering is ready to receive the package. Advisors set their own client fee and can currently run white-label assignments at $495 per named client during the founding cohort.

No. You can start with public context, notes, and partial evidence. The workflow highlights what evidence gaps matter most before major investment.

It can gather and organize evidence from a company URL, documents, spreadsheets, and optional employee-approved workflow summaries. It compares that evidence with a broad library of opportunity patterns and ranks the ideas the evidence supports. Machine-drafted facts remain suggestions until a person reviews them.

No. Advisors, investment and portfolio teams, transformation teams, and functional leaders can all use the same evidence-first workflow for their own decisions.

General prompting can brainstorm ideas. Use Case Foundry keeps evidence, scores, and the order of work connected so decisions are easier to defend.

Yes. Teams can use redacted or aggregated evidence and control what facts are accepted before recommendations change.

No. Capture is opt-in per employee, records no screenshots, keystrokes, or file contents, and summarizes activity locally on the employee's machine. Only a redacted summary of recurring workflow patterns is shared, denylists exclude chosen apps, sites, and folders, and nothing enters the company model without human review.

A ranked roadmap that separates fund-now ideas, required preparation, evidence gaps, and rejected generic ideas. Selected opportunities can continue into adoption plans, decision briefs, simulations, agent designs, tests, supervised pilot plans, production-handoff packages, and downloadable starter code.

Not as a universal one-click service. Most opportunities end in a reviewed engineering package and starter code. Tested customer-pilot paths exist for support copilot, evidence assembly, and financial reconciliation, each with its own limits and customer acceptance testing. The customer owns the project, access, credentials, and go-live decision.

The initial analysis and included deployment demos do not need an AI model key. Deeper use-case writing, adoption planning, and live use of an exported AI-backed agent require an approved OpenAI-compatible service. Without a key, the product can prepare prompts for use elsewhere.

A finished example opens in minutes. A real company assessment usually involves several people and working sessions: start with a URL or files, review the first analysis, then close the most important evidence gaps before preparing delivery materials.

Yes. The same server runs locally or in Docker and has documented cloud deployment options. Optional sign-in keeps saved company maps private, and AI-service keys stay on the server. Assisted work can run entirely inside your network.

Yes. Scoring dimensions and rationale are visible so stakeholders can review and challenge assumptions directly.