US 12,452,565 B1
Interpretation and transformation of unstructured human input to enable end-to-end service visibility
Daniel W. Rose, Red Bank, NJ (US); Timothy Wayne Masse, Spring Hill, TN (US); and Frank M. Massa, Red Bank, NJ (US)
Assigned to LIGHTRIVER TECHNOLOGIES, INC., Concord, CA (US)
Filed by LIGHTRIVER TECHNOLOGIES, INC., Concord, CA (US)
Filed on Jul. 1, 2025, as Appl. No. 19/257,042.
Application 19/257,042 is a continuation of application No. 19/217,611, filed on May 23, 2025.
Int. Cl. H04Q 11/00 (2006.01); H04L 41/16 (2022.01)
CPC H04Q 11/0062 (2013.01) [H04L 41/16 (2013.01); H04Q 2011/0073 (2013.01); H04Q 2011/0083 (2013.01); H04Q 2213/13343 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A computer implemented method for determining an end-to-end service path, comprising:
(a) assembling a prompt comprising (i) a plurality of example circuit names, (ii) for each example circuit name in the plurality of example circuit names, data points corresponding to circuit endpoints of a service corresponding to a respective example circuit name, and (iii) a circuit name for interpretation;
(b) sending the assembled prompt to a large language model (LLM) to determine first data points of a first endpoint and second data points of a second endpoint, wherein the first endpoint and the second endpoint belong to a service path corresponding to the circuit name; and
(c) deriving using the first data points, a first neighbor connection to be created corresponding to the first endpoint, and using the second data points, a second neighbor connection to be created corresponding to the second endpoint,
(d) wherein the first neighbor connection to be created corresponding to the first endpoint and the second neighbor connection to be created corresponding to the second endpoint enable determination of the service path.