AI-assisted Development: Providing Performance Feedback for API Programs
summary
Claims development-time forecasts of throughput and latency for integration programs by combining per-endpoint models with a system model, using observed latency and requests-in-flight as training features. The pipeline instruments deployed APIs via the observability stack, stores samples in a queue, fits per-endpoint models (analytical Universal Scalability Law or ML like XGBoost), extracts API calls from the AST, and solves a closed queuing network for the whole flow. Demo in Choreo IDE renders concurrency curves and per-call breakdowns (e.g., 20 ms and 35 ms) and matches production within observed-data limits.
AI-written summary · about
chapters (11)
- 0:00Introduction
- 0:25Main Use Case
- 1:57Why Use APIs
- 6:43Why Performance
- 8:33Benefits
- 10:08Performance Metrics
- 10:44Latency
- 11:58Objective
- 14:30Data
- 17:27Application Development
- 18:25Results
related talks
47mNext generation of API testing : Unleashing the power of AI
22mLeading the future: AI and APIs, a powerful duo
51mAI-Driven Strategies for Scalable & Resilient Performance Engg.
27mAI-Enhanced API Documentation [...]
13mApplying Azure Machine Learning service to drive success of API programs
27mAI as a Service Scaling AI with Ease and Innovation
25mAPIOps: automating API operations for speed and quality at scale
22mAPI Product Management with an AI tools & mindset