Experience: 7–10 years

Own the testing and evaluation approach for Dailoqa’s Broccoli platform, working closely with Capability Leads, AI Architects and platform engineers.

1. Define agentic testing standards: Establish test strategies covering individual agents, tools and end-to-end workflows, including multi-agent coordination, state management, failure recovery and human approvals.
2. Build automated evaluation capabilities: Develop reusable evaluation frameworks, representative datasets and regression suites to assess task completion, factual accuracy, groundedness, tool selection and adherence to instructions.
3. Test safety, security and controls: Validate agent permissions, data boundaries, audit trails and escalation paths; test prompt injection,



sensitive-data leakage, unauthorised actions and unsafe tool use.
4. Embed quality gates into delivery: Integrate automated tests and evaluations into CI/CD, defining release thresholds and comparing results across changes to models, prompts, tools and platform capabilities.
5. Validate production performance and drive improvements: Test latency, cost, scalability and reliability; reproduce live failures and turn findings into regression tests. Partner with Capability Leads to resolve defects and strengthen platform quality.

📌 Sr. Agentic AI Testing Engineer (Noida)
🏢 Dailoqa
📍 Noida