Pramana (working name) by VedAGI, Canadian workflow

Canada: in progress.

The Canadian workflow for Pramana (working name) is in progress and is not offered as a product. This page describes intended scope, not a live Canadian deployment.

The intended workflow applies source recalculation, labeled estimates, a named signer and a browser-verifiable receipt to life-sciences evidence packs. The US workflow is a production candidate. The figure-record version chain is being added and is not live today.

The name Pramana (working name) comes from Sanskrit, where it means proof: a means of valid knowledge.

Try it: Year-1 budget impact

Not checked
  • Application text$2.1M
  • Application text, French version for INESSS2,1 M$
  • pCPA briefing text$1.9M
  • Executive summary textabout $2M
Share of patients using the drug (uptake) 12%Estimate
Figure
Year-1 budget impact, public drug plans
Value
$2.1M (CAD)
Formula
treated_y1 × net_cost − displaced
Sources
BIA_model.xlsx, Results!C14
AI use
Formula suggested by AI, value worked out in code
Signed by
Named signer, market access

Example numbers only. This checks drafted text going out through Pramana (working name). Checking slides and decks against the number comes later and is being built. Pramana (working name) blocks the pack and people fix it. It never quietly edits a number.

The idea

The whole idea in one minute

Teams already build the model and already use AI to write the pack. The number then hides inside the writing, a person signs it anyway, and the other side cannot check it.

Pramana (working name) works each number out again from the source and marks guesses as guesses. It refuses to send until a person signs, and it gives the other side a receipt they can open in a browser. Same pack. Same model. The leak is closed.

  • Keepyour model
  • Keepyour writer, AI or human
  • Keepyour reviewers and signer
  • Addthe lock on the numbers
906 days

average wait from Health Canada approval to a first public listing1

27%

higher 3-year budget impact, median, after the agency re-ran company numbers2

+652 days

average extra wait after a “do not reimburse” recommendation1

The path

How a Canadian evidence pack gets made

A company has a drug approved by Health Canada and wants public drug plans to pay for it. The numbers travel a long way before anyone is covered.

01

The science is already done

Trials, prices and patient counts sit in spreadsheets and source files.

02

They already have a model

A cost-effectiveness model and a budget impact model. Canada's Drug Agency asks for models it can trace, with no hidden shortcuts 3.

03

The team builds the application

A reimbursement review application to Canada's Drug Agency, a French version for INESSS (Institut national d'excellence en santé et en services sociaux) in Quebec, and later materials for the pCPA (pan-Canadian Pharmaceutical Alliance) and the provinces.

04

The words get written The leak starts here

Often by an AI assistant now. The agency says any AI use must be declared, and the company stays accountable for every word 4.

05

Internal review and sign-off

Medical, regulatory and market access leads approve it and a named person signs. They trust the numbers match the model.

06

The agency reviews and re-runs

Reviewers at Canada's Drug Agency or INESSS often re-run the model and change the company's numbers 2.

07

Negotiate, then list

pCPA negotiates a price and signs a letter of intent, which is only an option to fund. Each province then signs its own listing agreement 6.

On average it takes 906 days from Health Canada approval to a first public drug plan listing 1. Any number that has to be explained or fixed adds months to that path.

The gap

Where it breaks: one number, many copies

The weak joint is not whether AI can write. It is that whoever writes each sentence can copy, round, translate, or quietly change the number inside it. In Canada the same number is reused in several documents and two languages.

One number, many copies The model holds $2.1M. Five documents copy it. One copy has drifted and one is rounded without a label. Readers see all of them. The model $2.1M Agency application $2.1M INESSS (French) 2,1 M$ pCPA materials $1.9M Changed in a rewrite.Nobody noticed. Provincial briefing $2.1M Executive summary ≈ $2M Rounded, with no label.Fact or guess? Readers Canada's DrugAgency INESSS pCPA Provinces
Example values only, for illustration.

Copies drift apart

One number lives in several documents and two languages. A hand edit, a translation or an AI rewrite changes one copy and not the others.

Guesses look like facts

An assumption about uptake reads like hard data because nothing marked it as an estimate.

Signers sign prose

The person signing cannot re-run every number from source. The agency then re-runs it and finds the gaps.

No trail when someone asks

When the agency or pCPA asks where a number came from, teams dig through emails and old spreadsheets.

Canadian rules already point this way

On AI use

Canada's Drug Agency asks companies to declare how AI was used, keep a capable human in the loop, and stay accountable for everything in the submission 4.

On models

The agency requires economic and budget impact models whose steps can be followed, with no hidden shortcuts, and its reviewers re-run them 3.

On promotion

When coverage claims are later reused in promotional pieces for health professionals, adding or changing them needs PAAB pre-clearance 5.

The industry already has models, writers and signers. It does not have a checkpoint that says: this number was worked out, labeled, signed, and can be checked by a stranger. That checkpoint is the gap. Pramana (working name) is only that checkpoint.

The checkpoint

What Pramana (working name) does

Keep the model. Keep the writer. Keep the reviewers. As the Canadian version is built, Pramana (working name) is meant to sit between “the words are drafted” and “the pack is signed and sent”, and check the numbers in the drafted text in both languages.

Where the Pramana (working name) checkpoint sits Sources, then model, then AI draft, then the Pramana (working name) checkpoint with four steps: re-run, label, lock, receipt. Then review and signing, then the readers. If a number is mismatched or unlabeled, nothing goes forward. Sources Model AI draft Pramana (working name) 1 Re-run 2 Label 3 Lock 4 Receipt Review and signer Canada's DrugAgency INESSS pCPA andprovinces Mismatched or unlabeled? Nothing moves forward.

Re-run

Each number is worked out again, in code, from the company's own sources. The AI may suggest a formula. It never writes the value.

Label estimates

If a number is an assumption, it is marked as one in the workbench (the on-screen work area), in the exported table, and in drafted text going out through Pramana (working name). There, it cannot pass as a measured fact.

Match the drafted text, or block

Numbers in the drafted text of the application, the French INESSS version, briefings and the summary are meant to match the locked number, or nothing is released. Slides in pCPA and provincial materials come later.

Hand over a receipt

What was worked out, from which sources, with which formula, signed by whom, and where AI was used. Reviewers open it in a browser, with no VedAGI account.

One number

The journey of one number

Follow a single number, the year-one budget impact for public drug plans, from the company's files to a reviewer's screen at Canada's Drug Agency.

  1. Start from the source

    Pramana (working name) reads the company's own files: trial data, prices, patient counts and the model. Nothing new is invented.

  2. AI may suggest, code decides

    The AI can propose how the number should be worked out. Pramana (working name) runs that formula itself, in code, on the real inputs.

  3. Check the drafted text

    The worked-out value is checked against the drafted text, in English and French. Any mismatch is flagged, not quietly accepted.

  4. Number or estimate

    Worked out from measured sources, it becomes a locked number. Built on an assumption, it stays a labeled estimate that everyone can see.

  5. A named person signs

    The signer approves worked-out numbers, not paragraphs. Pramana (working name) records who signed, when, and what role AI played.

  6. The receipt travels with the pack

    Value, formula, sources, signer and AI use, in one page any reviewer can open. It can help a company prepare the AI-use declaration the agency asks for. It is not a compliance check.

Receipt PR-0042
Figure
Year-1 budget impact, public drug plans
Value
$2.1M (CAD)
Formula
treated_y1 × net_cost − displaced
Sources
BIA_model.xlsx, Results!C14; price_list_2026.csv
Checked in
Drafted text of the application, INESSS (FR) version, pCPA briefing and summary: all numbers match
AI use
Formula suggested by AI; value worked out in code; human reviewed
Signed by
Named signer, 27 Sep 2026

Locked. Anyone can recheck this number.

Example receipt. Values are illustrative.

Value

What it is worth

Pramana (working name) does not make a product look better. It saves the hours spent chasing numbers and protects the launch from avoidable slips. Anything we estimated is marked as an estimate, the same rule Pramana (working name) applies to its own numbers.

Time spent tracing numbers, per pack

Today: checked by hand, about 3 review rounds
about 80 hours
With Pramana (working name), once the Canadian version is ready: application, French version and summary text checked by Pramana (working name); pCPA materials still checked by hand
about 20 hours, plus review of flagged items
up to about 60 hours savedper pack. Source: VedAGI's own estimate, worked out below, not a published study.Estimate

How we got this: About 40 key numbers, each repeated in about 4 places (agency application, French INESSS version, pCPA materials, summary), makes 160 spots. At about 10 minutes each by hand, that is about 27 hours per review round, times 3 rounds: about 80 hours. Pramana (working name) checks drafted text going out through it. Once the Canadian version is ready, that would cover the application text, the French version and the summary: 120 spots, about 20 hours per round, about 60 hours over 3 rounds. pCPA materials often include slides, so they (40 spots, about 20 hours) are counted as still checked by hand. The time to review what Pramana (working name) flags is not subtracted, so the saving is an upper bound. Re-running and locking the 40 numbers themselves is not counted, because this method has no assumption for it.

What a detour costs on Canada's path

906 days

average from Health Canada approval to a first public listing 1

  • After a “do not reimburse” recommendation+652 days on average 1
  • After a failed first price negotiation+490 days on average 1

Most detours come from real disagreements about value, which Pramana (working name) does not fix. It is meant to remove the avoidable ones in what goes out through Pramana (working name): numbers nobody can trace, drafted text that disagrees with the model, guesses read as facts.

A locked pack is cheaper than a corrected one
QuestionLocked before it leavesCorrected after someone else finds it
When the error is foundInside the company, before sign-offAfter the agency has already rerun your math
Who finds itPramana (working name) flags it automaticallyAgency reviewers, INESSS, or pCPA negotiators
What it costsA few hours to fix a flagged numberClarification rounds, a weaker negotiating position, or a resubmission
What you can say“Here is the receipt for the numbers we released.”“We will look into it and get back to you.”

We will measure, not guess. Until the Canadian version is ready, every time estimate here is only an estimate. Measured hours from real engagements will replace them.

Who it helps

Who it helps

Each person who touches the number gets something different.

  • The person who signs

    They approve worked-out numbers, not paragraphs. If a number cannot be locked, it does not go out.

  • The HEOR (health economics and outcomes research) and market access team

    They keep the model they trust and stop chasing numbers through the drafted text, in both languages.

  • The company that owns the product

    On a 906-day path, a number that has to be explained or resubmitted costs months and weakens its position at the pCPA table.

  • Canada's Drug Agency, INESSS, pCPA and the provinces

    They can check the receipt without buying anything. Trust moves from “we like this company” to “we can re-run this.”

  • Legal, regulatory and compliance

    A record of sources, formula, value, signer and AI use that can help answer questions from the agency. It is not a compliance check, and the agency does not endorse Pramana (working name).

Limits

What Pramana (working name) is not

Pramana (working name) does not:

  • make a drug look more cost-effective
  • replace the review by Canada's Drug Agency or INESSS
  • replace pCPA or a province's decision
  • write the submission
  • remove the human from the process
  • give legal or regulatory advice

If the model is wrong, the lock will faithfully lock the wrong math. The benefit is a trustworthy release, not a better economic answer.

Start

Follow the Canadian version

The Canadian version is in progress and is not offered as a product yet. If you prepare submissions to Canada's Drug Agency or INESSS and want to follow it as it is built, talk to us. For work on a live dossier today, start with the United States AMCP version.

Sources

  1. PMPRB, NPDUIS: Factors delaying the public listing of drugs in Canada (2025), 160 drugs first listed 2018 to 2023. Link
  2. Health technology assessment reports for non-oncology medications in Canada, 2018 to 2022 (PMC11499573). Link
  3. Canada's Drug Agency, Pharmaceutical Reviews Update 23 and Procedures for Reimbursement Reviews. Link
  4. Canada's Drug Agency, Position Statement on the Use of AI in the Generation and Reporting of Evidence (April 2025). Link
  5. PAAB, Advisory on Provincial Formulary Coverage Statements. Link
  6. pCPA, Standard negotiation process; C.D. Howe Institute, Faster deals or faster access? (2025). Link

Time and money figures marked as estimates are illustrative and will be replaced by measured results from the first engagements. Nothing on this page is legal or regulatory advice. Looking for the other market? Read the United States edition.

Get in Touch

If you are about to send a table whose numbers you cannot re-run, talk to us.

VedAGI. Intelligence for Humanity.