US 12,236,202 B1
Adaptation to detected fluctuations in outputs from artificial intelligence models over time
William DeWeese, Cocoa, FL (US); and Erich Stuntebeck, Johns Creek, GA (US)
Assigned to Airia LLC, Alpharetta, GA (US)
Filed by Airia LLC, Alpharetta, GA (US)
Filed on Jul. 26, 2024, as Appl. No. 18/786,402.
Claims priority of provisional application 63/658,434, filed on Jun. 10, 2024.
Claims priority of provisional application 63/650,487, filed on May 22, 2024.
Claims priority of provisional application 63/648,162, filed on May 15, 2024.
Int. Cl. G06F 17/00 (2019.01); G06F 40/30 (2020.01); G06F 40/40 (2020.01)
CPC G06F 40/40 (2020.01) [G06F 40/30 (2020.01)] 20 Claims
OG exemplary drawing
 
1. A method for maintaining consistent results in an artificial intelligence (“AI”) pipeline, comprising:
submitting, by a prompt engine, a first sequence of content queries to a pipeline engine for the AI pipeline, wherein the AI pipeline includes a prompt package and a language model, and wherein the prompt package and at least a portion of the first sequence of content queries are inputs to the language model;
receiving first results from the AI pipeline in response to the first sequence;
in an instance subsequent to receiving the results from the AI pipeline, submitting, by the prompt engine, the first sequence of content queries to the pipeline;
receiving second results from the AI pipeline;
vectorizing the first and second results;
semantically comparing the vectorized first results and the vectorized second results;
identifying a semantic divergence by determining that the first results and the second results differ semantically more than a predetermined threshold amount; and
in response to the determination, performing a corrective action.