Panel Discussion: Agentic Engineering and Quality in Healthcare
summary
Panelists report concrete results: a $250K consultancy scope is replaced by augmented teams, test authoring in the IDE and first‑pass code review save hours, and iterative self‑improvement of LLM‑written suites mutation score plateaued at 53% until given concrete failures — show it the breaks and it jumps to 90%. Validation shifts to protocols and continuous auditing, since pinning the model is a dependency with a clock; operational fixes favor reversibility and one flag per flow with a removal date.
AI-written summary · about
chapters (52)
- 0:00Agentic Engineering and Quality in Healthcare
- 0:35Panel Intros - Chad Ullman, Rohit Jaiswal, and Mark Kroh
- 3:39Beyond the Hype: Where Agentic AI Actually Moves the Needle
- 4:09Conservative by Nature - and Why That Wasn't Doing Nothing
- 4:41Replacing a $250K Consultancy Engagement With Augmented Teams
- 5:11AI as a Superpower, Not a Super Replacement
- 5:41Test Authoring in the IDE and Automated First-Pass Code Review
- 6:12The Mutation Score Experiment: Self-Improvement Plateaus at 53
- 6:43Show It the Specific Breaks and the Score Jumps to 90
- 7:13LLMs Aren't Short on Capability - They're Short on Feedback
- 7:44Amplifying the Tester and Tracing Coverage Back to Requirements
- 9:47Point-in-Time Validation Doesn't Work for Systems That Adapt
- 10:18Sign Off on the Protocol, Not the Artifact
- 10:48Borrowing the FDA's Predetermined Change Control Plan
- 11:19Your Eval Suite Becomes Your Regression Suite
- 11:49Pinning the Model Is a Dependency With a Clock on It
- 12:20From One-Time Sign-Off to Continuous Auditing
- 13:51Traceability When Something Fails Is the Part That Stays
- 14:53Understanding the Contract: Requirements Become More Important, Not Less
- 15:55What "Safe to Ship" Means in a Highly Regulated Domain
- 16:59The Duplicate Claim That Used to Take Months to Surface
- 18:01Testing Distributed Across the Pipeline - Everyone Owns Quality
- 19:32The Engineering Director's Answer: How Fast Can I Un-Ship?
- 20:34Investing in Reversibility: One Flag Per Flow With a Removal Date
- 22:06Synthetic Data vs Real World Edge Cases
- 22:37De-Identified Production Data and Querying It With AI
- 23:38Data Mining Instead of Synthetic Generation
- 24:08The Structural Problem: Privacy Deletes the Tail of the Distribution
- 24:38Stop Measuring Fidelity - Measure the Synthetic-to-Real Delta
- 25:08"Vintage": Every Synthetic Patient Is Born Today
- 25:38You Can't Test Your Way to Edge Case Coverage
- 26:10Feeding Production Incident Tickets Back to Generate Edge Cases
- 27:40Why De-Identified Data Goes Stale and Breaks Your Automation
- 28:43Human in the Loop vs Human on the Loop
- 29:13Why Human Review at Agent Scale Degrades Into Rubber Stamping
- 29:43The AI Reviewer That Silently Stopped Reviewing
- 30:15Attribution and Giving the Reviewer a Signal, Not a Pile
- 30:45Three Deliberate Places for Humans: Before, During, After
- 31:16Never Bypass the Loop - Building Fundamentals in New Testers
- 33:18The Future of the Automation Engineer
- 34:18Becoming a Communicator Instead of a Coder
- 35:18Four Priorities: Oracle Design, Eval Design, Reading a Suite Critically
- 36:19The Most Underrated: Test Data Generation as Engineering
- 36:51Business Domain Understanding as the Real Moat
- 37:22One Strategic Objective to Take Back Tomorrow
- 38:25Invest in Your People as Much as Your Technology
- 38:56Adopt the Agents but Own the Context
- 39:27Know Which of Your Controls Are Mechanisms and Which Are Just Sentences
- 39:57The Replit Incident: A Code Freeze That Was Text, Not a Permission
- 40:28Go Verify Every Required Check Actually Runs, Not Just Passes
- 42:01Q&A: HIPAA Compliance and Agent Workflows
- 44:05Q&A: Are We Creating an Accountability Gap?
related talks
37mBest Practices for Accelerating Delivery
50mPanel: Beyond The Vibe: Code Quality First
59mPanel Discussion: Strategies for Scaling AI in Quality Engineering
01h 05mPanel Discussion: Building Confidence in AI-Driven Testing Decisions
20mFireside Chat
47mPanel Discussion: QE Strategic Shift: What's Changing with AI, Automation, and Speed?
44mPanel Discussion | The AI Revolution in Testing: Opportunities and Challenges
01h 05mPanel Discussion - The Future of Testing: Impact of AI in Quality Assurance and Beyond
