Find your most expensive lines of code with Amazon CodeGuru
summary
Amazon announces Amazon CodeGuru Reviewer and Profiler as ML-powered tools that find expensive lines of code, flag hard‑coded AWS credentials, missing DynamoDB pagination, path‑traversal and resource leaks, and add new security detectors for top OAuth issues; Reviewer supports Java 8–11 and Python 3.6–3.9, Profiler adds heap summaries and continuous CPU/memory profiling. Atlassian replaces custom tooling with the JVM internal profiler, fixes EnumMap -> ConcurrentHashMap contention and cuts an object hotspot from 16% to 5%, reporting ~60% cost reduction in one environment; 90‑day free trial and 30,000 lines free/month for repo analysis.
AI-written summary · about
chapters (17)
- 0:00Intro
- 0:29Add Amazon CodeGuru to your typical software workflow
- 2:10[GA] Security Detectors by Reviewer
- 4:26Example: AWS API security best practices
- 9:20[Public Preview] Python support by Reviewer
- 11:32Example: AWS Python API best practice by Reviewer
- 14:38Introduction
- 15:36Context and background
- 17:13Adopting Amazon CodeGuru Profiler
- 18:18Use case 1: Improve CPU utilization and latency
- 20:01Use case 2: Hot spot view
- 21:47Performance journey continues ... memory
- 23:13[GA] Memory profiling
- 23:49CodeGuru Profiler memory use cases
- 27:20[Public Preview] Python profiling (CPU)
- 28:16Python profiling visualizations
- 28:54Summary of new features
related talks
22mThe future of software development, powered by machine learning
28mA Modern Application Builder’s Guide to Testing
28mA Modern Application Builder’s Guide to Testing
17mTransform your bug bash into an AWS BugBust
26mAWS Developer Innovation Day opening keynote
45mCracking the Code Review
45mCracking the Code Review
48mJava's Hidden Gems: Tools and Libraries