Testing at the Speed of Thought: QA in the Era of GenAI apps
summary
Claims that QA can run “100x faster” and compress two‑week sprints to hours, supported by a demo of an open‑source agent called Hercules open source testing agent that you can pip install Hercules. It spawns browsers with Playwright, plans steps via a supervisor pattern (planner + subagents), runs UI/API/accessibility/security/mobile checks, records screenshots, network logs, Playwright traces, video and emits HTML plus JUnit XML; uses multimodal LLMs, guardrails, and recommends larger models for repeatability.
AI-written summary · about
chapters (19)
- 0:00Introduction
- 1:26Sprint cycles and agile delivery
- 2:47Cost of poor quality and QA challenges
- 4:11Dev agents accelerating delivery
- 5:12Introducing Hercules testing agent
- 5:55Live demo: Hercules testing e-commerce
- 9:51Architecture of Hercules testing agent
- 12:40Future of agile testing with AI agents
- 13:22Rethinking sprint cycles with 100x teams
- 14:24Q&A and book giveaway
- 16:27Guardrails and hallucination prevention
- 18:46Natural language test case writing
- 20:53Integrating test cases with user stories
- 21:36Open-source vs SaaS testing agents
- 21:57Tests supported by Hercules
- 23:53Test data injection and repeatability
- 24:59Hard vs soft test case writing
- 26:02Reviving ATDD with AI
- 27:25Closing remarks
related talks
48mBeyond Bugs: How AI Turns QA into Full-Cycle Engineers
31mScaling Quality in a Decision Intelligence Platform: The Agentic QA Playbook
18mReinventing App Testing with Agentic Apps
29mQA vs AI: Friends or Foes?
30mAgentic QA vs LLM Dev: A Live AI Showdown in the DevOps Arena
45mSpyros Katopodis: How You Should Be Using Agile and AI For Testing And Quality Assurance
46mQA to QE: Scaling Quality with Ownership, Code, and Culture
22mBeyond Bugs: How AI turns QA into Full-Circle Engineers