Frequently asked questions
Premise AI turns a product idea into a tested business case and a Fund, Pilot, Kill, or Park recommendation. These answers cover what it is, how the numbers are derived, what is stored, and what it costs.
What is Premise AI?
Premise AI turns a product idea into a tested business case. It weighs market entry potential, the path to revenue, the resource demands, and the investment return, then places the result on a decision map so the call is clear: fund, pilot, kill, or park.
Who is Premise AI for?
Leaders who decide where people, budget, and time go: product and strategy executives, founders weighing the next move, and the teams that have to defend a resourcing decision to the people holding the purse.
How is this different from a general-purpose AI assistant?
A general-purpose assistant will draft a case when asked, but each answer is shaped by how the question was posed. Premise AI applies one evidence model to every idea the same way: market entry potential, the path to revenue, resource demands, and investment return, each figure carrying the assumptions behind it, placed on a fund, pilot, kill, or park map. The result is comparable across ideas and built to be defended, not a one-off narrative.
Is Premise AI affiliated with an AI provider?
No. Premise AI is an independent tool. A live analysis runs on the operator's own account with the AI provider they choose (Anthropic, OpenAI, Azure OpenAI, or a custom endpoint); it is not built, endorsed, or operated by any of them.
How are the numbers derived?
Every figure comes from the idea as described, verifiable external sources when available, and a consistent set of visible assumptions about adoption, pricing, cost, and timing. Estimates remain labeled as estimates, and each metric names the evidence that would most change the read.
Can the assumptions be changed?
Yes. Adoption, pricing, ramp, and cost assumptions are visible in each module and can be adjusted. The economics and recommendation recompute to show how sensitive the call is to the inputs most in doubt.
How does Premise AI size a market?
It works from the top down and the bottom up: total addressable market narrowed to a realistic beachhead by segment, geography, and willingness to pay, then checked against an adoption and pricing build. Where the two disagree, the gap and its cause are shown rather than averaged away.
How does Premise AI measure AI feature quality?
An AI feature is scored on the things that decide whether it ships and holds: data readiness, the quality gate it must clear, the cost to serve, and the risk of getting it wrong. Each is rated with the evidence behind the rating, not a single black-box grade.
Why are competitors anonymized?
Comparisons describe rival positions by their strategy and economics rather than by name, so the analysis stays about the structural question and avoids implying claims about specific named companies that it cannot prove.
What is stored, and where?
The API key and the reports generated are kept in the browser. Nothing is sent to a Premise AI server. The only outbound request goes to the chosen AI provider (Anthropic, OpenAI, Azure OpenAI, or a custom endpoint) to run the requested analysis.
Is Premise AI free, and how much does it cost?
The example report and the full interface are free to explore. A live analysis runs on the operator's own provider key (Anthropic, OpenAI, Azure OpenAI, or a custom endpoint) and costs only that provider's API usage, with no fee added on top.
Is an account or API key required?
No account is required to read the example report. A live analysis needs an API key for the chosen provider (Anthropic, OpenAI, Azure OpenAI, or a custom endpoint), which stays in the browser and is sent only to that provider to produce the report.
Can the results be exported?
Yes. Any report can be saved as a PDF for circulation, and it is structured so it can be reviewed module by module with the relevant teams.
Start with the next idea asking for resources.
Build the case and see where it lands on the board.