Industries
Pramana (working name) is built for pharmaceutical, biotech, and medical technology teams whose figures go to payers and health technology assessment reviewers. In that work, every number in the pack has to be recomputed from source, defended under questioning, and signed by a person. Serlyx, VedAGI's other product, is for insurance underwriters.
01 · Primary focus
Pramana (working name) is built for the teams that put figures in front of payers.
Manufacturers send payers the numbers that coverage and price decisions rest on: budget impact, eligible population, uptake, cost per patient, and per member per month (PMPM) cost. Those numbers start in an economic model, then get copied into dossiers, value decks, and submissions. AI is moving fast into this work, and large language models (LLMs) can produce figures that look right but cannot be traced to a source. Pramana recomputes each figure from your own model, labels every estimate, and locks the figures until a named person signs. Today that runs on the US AMCP Section 4 budget-impact table, the production candidate profile. Pramana is also adding a record of each figure across versions, so a revision released through Pramana must be a new version of a figure it already holds. That part is being built and is not live yet. AI may plan the work. Code computes the numbers. A human signs.
The production candidate profile. AMCP dossiers for payers and P&T (pharmacy and therapeutics) committees, starting with the Section 4 budget-impact table.
Not offered as a product yet. A profile for submissions to CDA-AMC (Canada's Drug Agency) and INESSS in Quebec, negotiated through the pCPA (pan-Canadian Pharmaceutical Alliance), where the same figure has to hold at every step.
02
The people on the other side of the pack.
Payer pharmacists, P&T committees, HTA reviewers, and negotiators have to decide whether to trust a number they did not build. Every figure Pramana (working name) locks comes with a receipt that says what was computed, from which sources, under which formula, and who signed it. The receiver can check that receipt in a browser, without installing anything, and without taking the number on faith. As the figure record is built, the same receipt will also show whether a figure is a new version of one sent before, and say plainly when that chain is missing.
03 · Serlyx
The people who decide whether to insure AI risk.
Serlyx helps underwriters assess AI risk and E&O (errors and omissions) exposure for companies that build or use AI. It is for E&O and technology-liability underwriters at insurers and MGAs (managing general agents). It produces a structured risk assessment and a binder-ready summary, labels every estimate, and leaves the decision to a human underwriter.
04
Life sciences comes first. The same lock applies wherever a figure leaves one organization and another organization has to rely on it. These are the areas we expect to extend Pramana into after life sciences:
Service-line, capacity, and reimbursement figures that go to boards and funders.
Figures in funding and policy submissions that have to be defended in public.
We work with a few market access, HEOR, and evidence teams at a time. Start with one table.
VedAGI. Intelligence for Humanity.