Understand context
The agent analyzes webpage content to understand the UI and system context it is operating in.
Case Study
Agent-based quality and governance
for an enterprise AI security platform.
Overview
Acuvity is an enterprise AI security and governance company that helps organizations discover, govern, and secure AI use across employees, applications, and agents. Its platform gives enterprises visibility and control over AI interactions while helping protect sensitive data and support responsible adoption.
In February 2026, Proofpoint acquired Acuvity, extending Proofpoint's platform with AI-native visibility, governance, and runtime protection for AI- and agent-driven workflows.
The Challenge
Acuvity co-founder and CEO Satyam Sinha met profiq CEO Jiri Manda at the Plug and Play Silicon Valley Summit in Sunnyvale. Their conversation quickly focused on profiq's research in AI-powered quality engineering and its potential fit for Acuvity's platform.
Acuvity needed an automated validation solution for enterprise-scale AI environments—one that could test complex platform interactions and AI behavior with minimal human intervention. The solution also had to evolve with the platform, work across AI providers, and support the security, reliability, and traceability expectations of large organizations.
Our Solution
profiq designed and implemented an AI agent-based validation framework that acts as a runtime quality and governance layer for Acuvity's platform. It automates the orchestration of multi-step validation scenarios, interacts dynamically with AI systems and external providers, continuously verifies outputs and behaviors, and scales testing across multiple model providers — simulating real-world enterprise usage and ensuring platform reliability under evolving conditions.
Project approach
Plan validation
It dynamically plans the multi-step validation scenarios needed to exercise platform interactions and AI capabilities.
Execute actions
It executes actions across systems and external AI providers, driving the browser through Playwright.
Evaluate outcomes
It evaluates results with LLM-based reasoning, making assertions through question-answering to confirm correct behavior.
Technologies and Tools
The framework is built in Python, with Pytest as the primary test runner and Playwright for browser automation. Its AI layer uses OpenAI GPT models, while the team also evaluates emerging capabilities such as Claude Computer Use and Microsoft OmniParser. Jenkins orchestrates the automation pipeline across Windows virtual machines and GUI-enabled Docker containers, and Allure provides test reporting and dashboards.
Benefits
Confidence in AI behavior
Increased confidence in AI system behavior in enterprise environments.
Reduced deployment risk
Reduced risk of deploying unreliable AI workflows.
Faster provider onboarding
Faster onboarding of new AI providers.
Scalable validation
Improved scalability of platform validation across multiple model providers.
Continuous evolution
Enabled continuous evolution of the AI platform without loss of quality.
Client testimonial
“Partnering with profiq has been a transformational experience. Their technical expertise in AI and enterprise solutions allowed us to bring a secure, scalable AI platform to market faster than we imagined. This collaboration has been instrumental in helping us offer a safe, innovative product for enterprises using generative AI responsibly.”
Satyam Sinha
Co-founder and CEO, AcuvityConclusion
Acuvity needed more than conventional test automation. It needed a validation approach that could keep pace with a fast-evolving enterprise AI platform.
profiq delivered an agent-based framework that orchestrates realistic scenarios, validates behavior across AI providers, and preserves traceability as the platform changes. The result is an extensible quality layer that supports reliable development and helps Acuvity evolve its AI security platform with confidence.