Course Overview
Most testers already use AI, but mostly as a chat window: paste a story, copy test cases out, fix the generic results by hand. Teams that are ahead have moved to agents that write, run, heal and triage tests inside the IDE, the browser and the CI pipeline. This course takes you there with the tools those teams use, one real QA problem per week, and leaves you with an AI-augmented Playwright framework you built yourself.
Claude Code, Cursor and Copilot agent mode writing whole Playwright suites from your project rules, and how to catch what they get wrong.
Playwright's own Planner, Generator and Healer agents, driven through Playwright MCP against a real app.
Build a LangGraph agent that reads Jira stories, writes and runs tests, and root-causes failures.
Run agents in GitHub Actions with approval gates, safe secrets, cost control and measured time saved.
From Chat to Agents: Level 1 → 5
Where are you today? Most testers are at Level 1 or 2. This course takes you to Level 5.
Paste a story into ChatGPT, copy the test cases out, fix them by hand.
Copilot autocompletes lines of test code as you type.
Claude Code or Cursor writes, runs and fixes a whole suite from your project rules.
Playwright MCP and Playwright Test Agents explore, plan, generate and heal tests in a live browser.
Your own QA agent triages CI failures and drafts Jira bugs for your approval.
Your Future Role
You finish this track ready for the QA roles teams are hiring for now: testers who ship faster with AI and know when not to trust it.
Uses AI agents across test design, automation and triage to cut cycle time without cutting quality.
Builds and maintains automation frameworks with AI coding assistants, MCP and self-healing built in.
Decides where AI fits in the team's QA process, sets review rules, and measures the real time saved.
- ✓ Build a QA prompt library that produces structured, importable test cases and data
- ✓ Generate a Playwright UI and API suite with Claude Code, Cursor or Copilot agent mode
- ✓ Spot the common flaws in AI-written tests: weak assertions, invented locators, false passes
- ✓ Plan, generate and heal tests with Playwright Test Agents and Playwright MCP
- ✓ Build a LangGraph QA agent that triages CI failures and drafts Jira bugs, with human approval
- ✓ Show an AI-augmented test framework on GitHub as your capstone
Syllabus Breakdown
6 Sessions • Saturdays • 2.5 Hrs/SessionPrompting That Works for QA
"AI gives me generic test cases." Role and context prompts, few-shot with your own test cases, structured output (JSON/CSV) ready to import into test management tools, and test data with Faker + LLM.
Tools: ChatGPT, Claude, Gemini • You build: a reusable QA prompt library
AI-Assisted Test Design
"Test design takes days per feature." Story to scenarios, boundary, negative and edge cases, spotting ambiguous requirements, risk-based prioritisation, and grounding AI on your own product docs.
Tools: Claude Projects, NotebookLM • You build: a reviewed test design for a real feature
Coding Agents for Test Automation
"Copilot only writes snippets." Agent mode in Claude Code, Cursor and Copilot; project rules (CLAUDE.md, Cursor rules); generating Playwright UI and API suites and page objects; and catching what AI gets wrong: weak assertions, invented locators, false passes.
Tools: Claude Code, Cursor, GitHub Copilot, Playwright • You build: an AI-generated suite you've hardened
Agents That Drive the Browser
"Locators break every sprint." Playwright MCP, Playwright Test Agents (Planner, Generator, Healer), and agent-led exploratory testing, including where the healer helps and where it hides real bugs.
Tools: Playwright MCP, Playwright Test Agents • You build: tests planned, generated and healed by agents
Build Your Own QA Agent
"Half my day is triaging red builds." The agent loop and tool calling, hands-on. An agent that reads a Jira story, writes tests, runs them and reports back, plus failure root-causing and flaky-test detection.
Tools: LangGraph, Jira MCP, Playwright • You build: a working triage agent
Agents in CI + Capstone
"How do I roll this out at work?" Agents in GitHub Actions with human-approval gates, auto-drafted Jira bugs, keeping secrets and data safe with coding agents, token cost, and measuring real time saved. Then present your capstone.
Tools: GitHub Actions, Jira • You build: an AI-augmented framework on GitHub
Real Tools, Real QA Workflows
You work hands-on with the tools QA teams are adopting now, and learn the workflow behind each one, so your skills carry over when the tools change.
ChatGPT, Claude, Gemini, NotebookLM
Claude Code, Cursor, GitHub Copilot agent mode
Playwright, Playwright MCP, Playwright Test Agents
LangGraph, MCP servers (Jira)
GitHub Actions, Jira
Hosted By
Rama Rajeswari
AI Practitioner and Entrepreneur. With 20+ years of industry experience and an IIT Guwahati AI certification, Rama specializes in enterprise Agentic systems, multi-agent workflows, and LLM evaluation pipelines. Spanning startups to Fortune-scale MNCs, her focus is on practical, battle-tested tool application over textbook theory.
Who Is This Course For?
Testers who want AI to speed up test design and automation, and to stay relevant as teams adopt it.
Framework builders who want AI coding assistants, MCP and self-healing in their stack.
Leads who need to roll out AI in the QA process with clear review rules and real results.