Bring us your hardest supply question.
Start with a few working sessions on realistic scenarios. No integration, no patient data and no production deployment. Go further only if the results justify it.
Three ways to work with us
Each level is scoped and agreed in writing. You choose how far to go, and you don't need IT, Quality and every vendor in the room to start.
Advisory
Review realistic scenarios and Veridose's outputs with us, and tell us where they're wrong.
- We ask
- Two or three structured sessions with your supply lead.
- You get
- Early prototype access, a say in what we build first, and a short findings memo.
- Data
- None. Synthetic scenarios only, so no IT approval is needed.
Design partner
Test Veridose on cases from your own trials, against success criteria we agree up front.
- We ask
- Representative or approved historical cases, and time to review the results.
- You get
- A confidential assessment of how Veridose handled your cases, and first option on a pilot.
- Data
- Historical or representative, under written data-use terms.
Pilot
One trial, read-only. Veridose recommends alongside your team, your team decides as usual, and we compare.
- We ask
- A named owner, plus scoped IT and Quality review of the connection.
- You get
- A joint evaluation report against your baseline, and negotiated commercial terms.
- Data
- Read-only access to agreed systems. Nothing is written back.
You can stop at any level. Advisory input is valuable on its own, and moving to the next level is a joint decision, not an obligation.
What we test together
We start with two workflows that matter most to lean supply teams, and judge Veridose against what your experts would have concluded.
Protocol amendment impact
A cohort, country or visit schedule changes. Veridose traces the effect on demand, inventory, packaging, labels, IRT settings and lead times, and shows where it isn't sure.
- Judged on
- Impacts missed, unsupported claims and review time, compared with your team's own analysis.
Impending site stock-out
Stock at a site is running low. Veridose forecasts the shortfall, uses only released stock eligible for that country, and ranks feasible resupply options.
- Judged on
- Forecast accuracy on past data, false alerts, and whether a planner would actually take the options.
We also test what happens when things go wrong: temperature excursions, attempts to infer treatment arms, conflicting documents, vendor records that disagree and outages. See the proof scenarios.
What we measure
We start with what can be seen directly in a pilot, measured against your current way of working.
Drug waste and days of delay avoided are estimated from these, with the assumptions stated, until longer use can measure them directly.
- Time to answerHours to assess an amendment or resolve a stock discrepancy.
- Forecast qualityAccuracy and range calibration against what actually happened.
- Records reconciledDisagreements between vendor systems caught, with their source.
- Review burdenFalse alerts, and the share of recommendations your team would act on.
How it runs after a pilot
Choose where Veridose runs. Data boundaries are agreed and documented before any data moves.
Managed
We host and operate Veridose on Azure in a region you choose, under agreed data-protection terms, including a BAA where the data requires one.
In your environment
Deployed in your own cloud account. Residency and processing boundaries are agreed up front, and every outside service, such as model hosting or support access, is documented.
Configured with your team
We tune models and scenarios to your trials together. Your team owns the assumptions; we own the models and their test evidence. Qualification for your intended use is done with your Quality team.
Who we're looking for
Teams with a real supply problem and someone who owns it. A large organization isn't required.
- Biotech sponsorsEmerging and mid-sized companies running clinical trials, with a named supply lead, outsourced CDMOs, depots and IRT, and past cases we can learn from.
- Specialist advisorsIRT, depot, CDMO, courier, Quality and security experts who can check one part of the design, without committing a whole team.
Agreed before any data is shared
Advisory work needs no data. Before anything moves at later levels, we put these in writing.
- Scope and exitWhat each side does, for how long, and how either side can stop.
- Data useWhat data is shared, how it may be used, including whether it may improve our models, and when it is deleted.
- OwnershipWhat each side brought, and who owns what we build together.
- PublicityNothing about your company, trials or results is published without your consent.
Start with one conversation
Tell us about your trial and the last supply problem that cost you time. We reply within a week.