US 12,141,560 B2
Hybrid-feedback driven transpiler system
Utkarsh Raj, Charlotte, NC (US); Paul Jacob Abernathy, Charlotte, NC (US); Vijaya Rudraraju, Charlotte, NC (US); and William Cruise, Charlotte, NC (US)
Assigned to Bank of America Corporation, Charlotte, NC (US)
Filed by Bank of America Corporation, Charlotte, NC (US)
Filed on Nov. 15, 2022, as Appl. No. 18/055,558.
Prior Publication US 2024/0160422 A1, May 16, 2024
Int. Cl. G06F 8/51 (2018.01)
CPC G06F 8/51 (2013.01) 20 Claims
OG exemplary drawing
 
1. A system comprising:
a source computing system comprising a data repository storing a source file associated with a source programming language;
a target computing system comprising configured to process operations in a target file associated with a target programming language, wherein the source programming language and the target programming language are different programming languages;
a transpilation platform, comprising:
at least one processor; and
memory storing computer-readable instructions that, when executed by the at least one processor, cause the transpilation platform to:
receive, from the source computing system via a network, the source file;
parse the source file to generate a first intermediate file corresponding to a structure of the source programming language;
transpile, by a machine-learning (ML) transpilation model, terms from the source programming language to corresponding terms in the target programming language;
generate, based on transpiled terms received from the ML transpilation model, a second intermediate file corresponding to a structure of the target programming language;
generate, based on the second intermediate file, a target file comprising instructions in the target programming language, wherein the target file causes the target computing system to perform actions performed by the source computing system based on the source file; and
train, by a language learning engine, the ML transpilation model, based on feedback received from the target computing system after performance of operations based on execution of the target file.