AI in Automation

What AI Can Do, What Testers Must Do

Partnering with AI in Automation

Ronak Ray

Vice President of QA & AI Strategy — Forbes

Speaker Notes: "AI is transforming how we test software — but transformation doesn't mean replacement. Today, I'll talk about where AI truly adds value and where testers remain irreplaceable."

Why This Talk, Why Now?

The New Reality:

  • AI is flooding software engineering.
  • Testing is under immense disruption.
  • 66% of DevOps teams are using or plan to use AI within 12 months.

Myth: AI replaces testers.

Truth: AI empowers testers who evolve.

Speaker Notes: "This isn't a future trend—it's happening right now. The central question isn't 'Will AI take my job?' The real question is 'How do I partner with AI to do my job better than ever before?' We've seen this before — recorders didn't replace musicians, spreadsheets didn't replace accountants. AI won't replace testers; it will amplify them."

The Grand Shift in Software Testing

⚙️ Manual ⚙️ Automation 🧠 AI-Augmented QA
Deterministic Probabilistic

Core Message:

We're moving from brittle scripts to intelligent, adaptive systems. As building gets faster, validation gets harder, and testing becomes the deterministic checkpoint in a non-deterministic world.

Speaker Notes: "For years, we've dealt with rigid, fragile scripts that demand constant maintenance. That era is ending. When AI writes code instantly, our bottleneck becomes validation. We're in the middle of a paradigm shift toward intelligent systems."

Part 1

What AI Can Do Well Today

The AI Arsenal: Augmenting the Testing Lifecycle

Intelligent Test Generation

Parse requirements from Jira/Figma to auto-generate test cases.

Self-Healing Automation

Adapts to UI changes in real-time, reducing script maintenance.

Predictive Defect Analysis

Analyzes historical data to predict bug-prone areas.

Advanced Visual Validation

Computer vision detects meaningful UI regressions.

Synthetic Test Data Generation

Creates realistic, privacy-safe data at scale.

Test Optimization

Identifies redundancies and optimizes scheduling.

Speaker Notes: "Think of AI as an arsenal of specialized tools. It can read our user stories and draft test cases. It tackles the nightmare of script maintenance. It acts like a fortune teller, pointing to risky code areas. It validates UIs with human-like nuance and generates complex test data without exposing sensitive information."

Deep Dive: Why Self-Healing is Your Best First Step

The Problem

Traditional scripts are fragile. A minor UI change breaks them, creating a huge maintenance burden.

The AI Solution

AI watches for changes during test runs. When an element's locator is modified, AI finds the new one and allows tests to continue without manual intervention.

The Business Case

  • Measurable reduction in engineering hours.
  • Immediate and undeniable ROI.
  • Builds crucial trust in AI.
  • Funds further investment.
Speaker Notes: "If you're wondering where to start, start here. The constant pain of flaky tests is universal. Self-healing provides a clear, quantifiable win. You can tell your manager, 'We saved 50 hours this month on test maintenance.' That success story is your gateway to exploring everything else AI offers."

What AI Can't (Yet) Do

Understand business intent

Judge what matters vs what just passes

Detect UX friction or emotional quality

Handle ambiguity, bias, or ethics

Provide strategic direction

Speaker Notes: "AI doesn't understand why something matters. It predicts patterns — it doesn't assess impact or risk. An AI can confirm a button works, but only a human can determine if the overall experience is intuitive and delightful."

Part 2

What Testers Must Own

The Tester's Evolving Mandate: From Executor to Strategist

Repetitive

Exploratory

Execution

Strategy

Functionality

User Experience

Your New Role: The AI "Manager" & "Trainer"

Speaker Notes: "Our value proposition changes dramatically. We're no longer just executors. We are quality strategists. We're the explorers who go off-script. We're the guardians of user experience. Think of yourself as managing a highly efficient AI assistant — you provide strategic direction and domain knowledge; AI provides the speed."

What Testers Must Still Own

Critical thinking & risk assessment

Contextual prioritization

Validating results against business goals

Ethical and governance oversight

Domain expertise & business context

Speaker Notes: "Testers are still the conscience of quality — the ones ensuring that what passes tests truly serves users and business needs. Your domain expertise is your superpower in the age of AI."

The High-Demand Skillset for the AI Era

Technical Skills

  • Prompt Engineering
  • Data Literacy

Strategic Skills

  • Domain Expertise
  • Critical Thinking

Essential Mindset

  • Adaptability
  • Continuous Learning
Speaker Notes: "To thrive, evolve your skills. You don't need to become a data scientist. Learn to 'talk' to AI through prompt engineering. Become comfortable analyzing data. But most importantly, amplify what you already have: deep business knowledge and critical thinking."

Part 3

How to Partner Effectively

The "Paired Testing" Model: Human + AI

AI is the "Driver"

  • Rapidly generates test scripts & data
  • Executes thousands of tests in parallel
  • Handles low-level technical details

Human is the "Navigator"

  • Sets strategic direction & testing goals
  • Observes AI execution & interprets results
  • Uses expertise to investigate anomalies
Speaker Notes: "How does this partnership work in practice? Let's call it 'Paired Testing.' The AI is the driver, with hands on the keyboard, executing at incredible speed. We are the navigators, providing strategic direction, watching results, and making the final call on quality. This is continuous co-evolution."

Partnership Framework

AI Can Do
Tester Must Do
Generate tests
Validate relevance
Write automation
Ensure maintainability
Find anomalies
Interpret impact
Suggest next steps
Choose what matters

Key Insight: AI is like a junior tester — fast but inexperienced. It needs human mentorship and validation.

Speaker Notes: "Think of AI as a co-pilot, not an autopilot. It can optimize routes, but you still need someone steering. AI provides speed; humans provide wisdom."

Case Study: Parallelization at Scale (Forbes)

The Challenge:

A growing regression suite was becoming a major bottleneck in our CI/CD pipeline, slowing down releases.

The AI-Assisted Strategy:

We paired human strategy (Navigator) with AI execution (Driver) to optimize, prioritize, and run tests at massive scale.

🚀 60%Faster Runs
📈 5.7%Higher Velocity
✅ 100%Coverage Kept
Speaker Notes: "Let me make this concrete with Forbes. Our test suite was becoming a monster. We applied the Paired Testing model. As navigators, we set the strategy. The AI did the heavy lifting—optimizing, prioritizing, and running tests at a scale we couldn't manage manually. The result? Higher-quality software, delivered faster."

Part 4

Your Actionable Framework

Getting Started on Your AI Journey

1

Assess & Identify

Find your biggest bottleneck.

2

Start a Pilot

Pick one high-impact area for a quick win.

3

Build Data Foundation

Centralize test results, defects, and logs.

4

Train & Scale

Upskill your team and expand.

Speaker Notes: "You can start this journey next week. Find your single biggest pain point. For many, that's test maintenance. Run a pilot on self-healing. Prove the value. That success gives you momentum to build your data foundation, train your teams, and scale from there."

AI Maturity Model

AI-Assisted

Tool-level (Copilot, Gemini)

AI-Augmented

Integrated into pipelines (CI/CD Anomaly Detection)

AI-Integrated

Autonomous systems (Self-healing orchestration)

Recommendation: Start at assisted, progress toward integrated — don't jump straight to autonomy.

Speaker Notes: "The most successful teams start small. Pick a repetitive workflow, experiment, measure, and expand. Don't aim for full autonomy on day one."

Metrics That Matter

AI Contribution Rate

Cycle Time Reduction

Coverage Confidence

AI Suggestion Accuracy

Engineering Hours Saved

Key Point: Success isn't just about test pass rate — it's about measuring AI-human synergy.

Speaker Notes: "Track the right things. Don't just measure pass/fail rates. Measure the partnership's effectiveness."

Pitfalls to Avoid

Blind trust in AI

Using AI as a shortcut for thinking

Poor data hygiene

Ignoring reproducibility

Skipping the pilot phase

Speaker Notes: "The fastest way to fail with AI is to stop thinking critically. Use it as an accelerator, not an excuse. Clean data is essential. Start small and prove value before scaling."

Redefining the Tester

From Bug Hunter Quality Strategist

Coach AI, don't compete Focus on system quality Elevate judgment & ethics Become an AI orchestrator
Speaker Notes: "The modern tester becomes an AI orchestrator — someone who ensures quality at scale, responsibly. We're moving up the value chain."

The New QA Formula

(AI Speed Tester Judgment) Quality at Scale

Your Path Forward:

🧭 Embrace augmentation 🔍 Redefine metrics 🤝 Build literacy & guardrails 🏁 Lead AI adoption
Speaker Notes: "This is the new partnership model — speed meets sense. That's the future of testing. We're not being replaced; we're being elevated."

Key Takeaways

AI is an Augment, Not a Replacement. The goal is a human-AI partnership.
Your Role Becomes More Strategic. Move from execution to strategy and user experience.
Partner Using "Paired Testing". Be the strategic "navigator" to AI's "driver."
Start Small, Prove Value, Scale. Begin with a focused pilot to build momentum.
Speaker Notes: "To wrap up: First, stop thinking about replacement and start thinking about augmentation. Second, recognize your value is shifting up the chain. Third, embrace the Paired Testing model. Finally, don't get overwhelmed. Start small, get a win, and build from there. The future of our profession is incredibly exciting."
"AI won't replace testers. But testers who embrace AI will replace those who don't."

✨ Thank You

Questions?

/in/ronakray

Speaker Notes: "Testing isn't disappearing — it's evolving. The ones who thrive will be those who partner intelligently with AI. Thank you, and I'm happy to take your questions."
1 / 20