Programme Launch Offer: Save 20% - Book Now

Track Talk, Th5

Multi Agent Cooperative Testing: Automation to Orchestration

Manisha Mittal

Apurva Singh

10:00 - 10:45 CEST, Thursday 18th June

You have invested years into building a sophisticated test automation framework. But as product evolved, the once-robust suite started showing its age.

We faced this exact problem. What was meant to accelerate testing became a heavy burden—duplicate tests piled up, flaky results eroded confidence, minor UI or API changes broke dozens of scripts, and feedback cycles slowed to a crawl.

Tests didn’t know what was relevant anymore. Execution was blind to context.

We didn’t need more automation at scale. We needed better orchestration and smarter execution.

During one intense meeting, a Gen-Z team member made a bold statement: “These tests should think, communicate, and self-organize.” That thought became the turning point.

What if different parts of the test framework could act like agents—autonomous, collaborative, and adaptive?

This vision led to Multi-Agent Cooperative Testing (MACT):

  • Specialized test agents that each have a role (e.g., Regression Agent, Sanity Agent, Risk-Detector, API Validator)

  • Local decision-making based on change deltas, risk scores, or telemetry

  • Inter-agent communication to avoid redundancy and build context

  • Continuous learning (e.g., “tests near this change area usually fail”)

MACT is a decentralized model where lightweight agents specialize in testing roles, collaborate in real-time, and adapt test behavior based on system change. It can sit on top of existing frameworks (like Selenium, Cypress, Playwright, REST Assured, etc.) and drive smarter execution.

In this presentation, we reveal how we navigated the crisis, emerged stronger with a robust framework powered by AI agents by sharing our journey from a rigid automation framework to a multi-agent testing ecosystem. We believe MACT represents not just evolution—but a leap forward in making “Testing at its Best” intelligent, distributed, and future-ready.

This talk will conclude with lessons learned, emphasizing resilience and how client challenges can be transformed into opportunities.