AI-Driven Quality Engineering Practices
summary
An AWS event‑driven pipeline combines CI/CD and CloudWatch logs with Lambda + SageMaker to cluster error messages and predict impact, running as an MVP for <$1,000/month and needing ~two hours/week ops. It vectorizes messages with a TF-IDF vectorizer (max_features=1000, ngram 1–2), clusters with DBSCAN clustering (eps≈0.3, cosine similarity), correlates timestamps across three core microservices, and surfaces predictions like “~1,200 users affected” so teams fix root causes instead of chasing 331 separate test failures.
AI-written summary · about
chapters (13)
- 0:00AI-driven quality engineering intro
- 1:29Reactive to predictive QE
- 2:30QE dashboards paradigm shift
- 4:20Alert fatigue with AI insights
- 7:07Budget-friendly AI test pipeline
- 8:35Event-driven real-time testing
- 12:53Error messages to ML features
- 16:26Failure pattern feature extraction
- 18:57Predicting business impact
- 21:07Future of autonomous testing
- 23:16AI integration blueprint
- 26:09Q&A AI in QA
- 31:23Closing Q&A
related talks
29mAI-Powered Anomaly Detection in Cloud Data Pipelines
12mLeveraging AWS and Kubernetes for Scalable AI/ML Workloads in the Cloud
44mBest practices for analytics and generative AI on AWS(ANT302)
26mServerless Architecture for Product Defect Detection
21mAI-driven optimization of CI/CD pipelines
23mBuilding Production-Ready Data Pipelines for AI/ML with Lambda Architecture
41mAlexander Gyulai: A Practical Guide to Transforming Quality Assurance with Intelligent Solutions
35mBring the power of machine learning to your developer operations