Treffer: Temporal Regression Forecasting: A Predictive Risk Nervous System for Modern CI/CD and AI-Assisted Applications

Title:
Temporal Regression Forecasting: A Predictive Risk Nervous System for Modern CI/CD and AI-Assisted Applications
Authors:
Satterfield, Adam
Publisher Information:
Zenodo
Publication Year:
2025
Collection:
Zenodo
Subject Terms:
devops, artificial intelligence, machine learning
Document Type:
Report report
Language:
unknown
Relation:
https://zenodo.org/records/16755887; oai:zenodo.org:16755887; https://doi.org/10.5281/zenodo.16755887
DOI:
10.5281/zenodo.16755887
Availability:
https://doi.org/10.5281/zenodo.16755887
https://zenodo.org/records/16755887
Rights:
Creative Commons Attribution 4.0 International ; cc-by-4.0 ; https://creativecommons.org/licenses/by/4.0/legalcode
Accession Number:
edsbas.8C99BB87
Database:
BASE

Weitere Informationen

**Temporal Regression Forecasting: A Predictive Framework for CI/CD Risk Analysis** This preprint introduces Temporal Regression Forecasting (TRF), a novel framework to predict software risks in CI/CD pipelines, addressing gaps in modern DevOps and QA. TRF combines two engines—Quality RIF (Ripple Impact Forecaster) for downstream effects and LRP (Latent Risk Predictor) for hidden vulnerabilities in unmodified code. Using features like code similarity (via AWS Bedrock/OpenAI embeddings), dependency graphs, and test health, TRF forecasts risks to guide testing and releases. Developed as Calliope Insight #1, it targets AI-assisted development’s challenges, such as emergent bugs. The paper includes Python examples for practical integration. Feedback welcome! Part of Project Calliope to advance predictive analytics in software delivery.