Skip to main content

Google Cloud Rapid Agent · Arize Phoenix · 2026

A denied appeal that gets better every time it loses.

Granum evolves insurance-appeal letters like an immune system. A naive draft matures into a champion that wins.

83%
overturned when filed
10%
ever appealed
12h
to write by hand
Aetna · Cardiac, 10 generations3 alive · 18 apoptosed
0.40.60.81.0gen 0g1g2g3g4g5g6g7g8g9
alive champion experimental apoptosed

One strategy starts at 0.40 and climbs to a 0.98 champion. The losers stay on the page, struck out.

how it works

It treats each appeal strategy like a living cell.

Every payer and diagnosis gets its own population. On each denial the strategies draft competing letters, a judge scores them against real overturned cases, and the losers are deleted from the registry for good. Winners branch.

Aetna · cardiac · seed
40%
appeal fitness, naive gen 0
after 10 generations
98%
appeal fitness, champion
strategies apoptosed
18
deleted from the registry, kept on record

Look inside a cell.

2 payers and diagnoses

Strategies that transfer across cells.

A champion matured against one payer is trialed against another payer's denials. The payer-specific citations don't carry over — negative selection runs on the appeal it writes in the new cell — but the matured strategy does. An edge is promoted only when it significantly lifts the target's naive baseline.

1 promoted transfer
no significant lift+0.19 lift · p=0.0240.98Aetna · Cardiac0.98United · Oncology
  • g7m0 (aetna) unitedrejected

    Scored 0.00 on united's denials versus its naive baseline 0.62 — not a significant lift (p=1.000, 5/5 rejected at negative selection). This strategy is clonally specific to its own cell.

  • g6m0 (united) aetnapromoted

    Scored 0.59 on aetna's denials versus its naive baseline 0.40 — a +0.19 lift, significant at p=0.024. The matured strategy writes better appeals in a payer it never trained on.