The registry, the reconstructions, and the open question

NextConsensus registers the probability that a named regulator, guideline body, or payer acts by a date, before the outcome is known, under a rule an independent reader applies afterwards. It also reconstructs how medical claims gain support, shed qualifiers, and spread across public sources, with a dated source trail for each. Whether that history improves the forecasts is the open question. Three ways in, below; the research program is the reading list.

What makes the work hard

The hard problem is reconstructing what was knowable at a date, defining what would settle a claim, and reviewing the record without hindsight.

1. Proposition identity across sources and time

   Authorities, actions, populations, deadlines, and resolution sources must stay comparable between the historical rebuild and the record registered in advance.

2. Temporal integrity

   Every record is pinned to an evidence cutoff. Later information cannot enter the registered state.

3. Resolution and review

   A correct-looking narrative is not enough. Outcomes need independent rules, non-events need structure, and any later claim about accuracy needs a denominator it can be traced to.

Three ways in

• Organizations. Embed Refract in your own pipeline, or co-build the healthcare layer on top of it. Send what you are building and the public sources it would read; we reply with whether the fit is real.
• Contributors. Refract is open source. Pull requests, issues, and docs are the best way to show how you work.
• Future hires. We are not hiring right now. Send a note to partner@nextconsensus.com with what you have built and we will reach out when a role opens; the first roles will be in applied machine learning, knowledge representation, clinical evidence science, and the infrastructure that keeps provenance intact.