The Role of API Testing in Agentic AI: Practical Perspectives

Join our newsletter to stay informed about Innovate QA’s virtual and in-person events, opportunities, news, and offers.
Subscribe to the Innovate QA Newsletter

Speaker: Dr. Srinivas Padmanabhuni | LinkedIn

Topic: The API Worked, the Agent Still Failed: The Role of API Testing in Agentic AI Testing, Practical Perspectives

Agenda:
🕘 9:00–9:05 AM PST — Welcome
🕘 9:05–9:55 AM PST — Dr. Srinivas Padmanabhuni | LinkedIn — The API Worked, the Agent Still Failed: The Role of API Testing in Agentic AI Testing, Practical Perspectives
🕘 9:55–10:00 AM PST — Q&A & Closing

Summary :
AI agents can fail even when the underlying APIs appear to work correctly. A wrong tool call, a retry that duplicates an action, a timeout treated as success, or a handover that loses context can lead to unexpected outcomes that remain hidden if we only evaluate the agent's final response.

In this session, Dr. Srinivas Padmanabhuni will explore how existing API testing practices can be applied to testing agentic AI systems. Using a hypothetical banking example, he will demonstrate how agent behaviour interacts with APIs and how different tests can identify failures before release.

The session will cover contract and schema validation for agent tools, permission and scope testing, retry and timeout handling, idempotency, lost acknowledgements, error-path testing, and verification of API call sequences.

Attendees will leave with a practical set of API-level tests they can incorporate into their pre-release checks for AI agents.

Takeaways :
What attendees will learn:

• How API testing practices can be applied to agentic AI systems.
• How to validate contracts and schemas for AI agent tools.
• How to test permission boundaries and ensure agents operate within their authorised scope.
• How to identify failures involving retries, timeouts, idempotency, and lost acknowledgements.
• How to test error handling and verify the sequence of API calls made by an agent.
• How to build a practical set of API-level tests for pre-release validation of AI agents.

🎙️ Speaker Opportunities at Our Meetups

Have expertise in QA, leadership, or tech skills you'd love to share? We're always looking for engaging speakers for our monthly meetups. This is a fantastic platform to showcase your knowledge, connect with the community, and inspire others. If you're interested submit your topic here — we’d love to hear from you!