Healthcare landscape rubric
Momentum and risk, from public record.
Scores an organisation as an investment or partnership candidate using sources anyone
can check. It first reads what the organisation makes and what its portfolio looks like, how much is on
the market and how much is still in development, from the same public record. A company is usually
both at once, so the two lenses are blended by that mix rather than chosen. It ends in a call and the
reasons for it.
- These are public proxies, not diligence. Every signal is a cheap observable standing in for
something expensive to measure properly.
- Absence of data is uncertainty, not safety. A signal with no record scores zero and is counted
as unknown, never as clean.
- Not applicable is not unknown. A company with nothing yet sold has no recalls, and a custom
device maker has no clearances because the law exempts it. Those rows are marked not applicable and
do not count as missing evidence.
- What it makes is inferred from the record and shown with the evidence used; it can be
overridden. Software and diagnostics do not separate cleanly in public data, so those two are best
set by hand.
- Stage is not asked, because a company has a portfolio, not a stage. Whether anything is
marketed and whether anything is in development are both read from the record, and the marketed and
development lenses are blended by that mix. A mature product company with a live pipeline is read
mostly, not wholly, as a product company.
- The weights are a judgement, and they change with the reading. Rows that carry no weight are
shown greyed rather than hidden, so the choice stays arguable.
- Name matching is by string. Subsidiaries, former names and legal entities that differ from the
trading name will be missed on every source.
This is a portfolio demonstration. The logic runs and the sources are live, but it was built to show an approach on a realistic problem, not deployed against a real one. Nothing here has been validated for clinical, commercial or investment decisions, and no result should be treated as advice.