Intro
Many AI tools for medicine promise more than they deliver. We help you choose and test them before you commit. We do not build or sell AI tools, so our advice depends only on the evidence.
Who it is for
- Clinics and hospital teams considering an AI tool.
- Health companies adding AI to a product.
- Research groups that want their AI results checked.
What we do
- Tool selection: compare candidate tools against your use case, your data and your workflow.
- Validation: check a vendor’s performance claims, and design a test on your own data before rollout.
- Study design: plan how to measure whether an AI tool helps, including endpoints and sample size.
- Model review: check a model’s training data, evaluation method and error analysis.
- Risk review: look for bias, data leakage and results that will not carry over to your setting.
Examples of questions we answer
- Does this imaging tool perform as well on our patients as in the vendor’s paper?
- Which of these three tools suits our workflow?
- Is the accuracy in this study real, or an effect of how the data was split?
- How many cases do we need to test before we trust the tool?
What you get
A written recommendation with the reasons, the evidence behind it and the checks we ran. For validation work, a test plan and a results report.
Process and timeline
- You tell us the decision you face.
- We reply with a written scope, price and delivery date.
- We do the work. The timeline depends on the task.
- We deliver the report.
Price
Fixed price for defined tasks, agreed in writing before work starts. Price on request.
Tell us which AI tool or model you are weighing.
FAQ
- Do you develop AI models?
No. We advise and validate. - Can you work under NDA?
Yes. We sign one before we see any data.