The project aims to evaluate the accuracy and reliability of these agents' outputs by reviewing security findings and assigning standardized labels based on defined guidelines. The scope also includes building a golden dataset of real security reviews and findings, each with an expert-verified correct answer, to benchmark agent performance and identify quality gaps

Candidate Requirements/Skills:

Security Engineer / Application Security Reviewer (primary):

- 4 to 8 years of Experience; secure code review; vulnerability identification and triage; exploitability reasoning; common classes such as injection, auth/authz, SSRF, insecure deserialization, secrets and misconfiguration; evidence-backed verdicts. GCP (Google Cloud Platform)/ Cloud Security Engineer (primary or complementary) :-GCP IAM, service accounts, networking, storage, workload identity, audit logging, cloud misconfigurations and least privilege; ability to judge cloud-specific exposure.

Calibration, rubric writing, reviewer alignment, adjudication tracking, benchmark versioning and structured quality reporting. Could be covered by one of the two engineers with client oversight.

📌 Freelancing Security Consultant (AI Evaluation Project) (India)
🏢 EduRun Group
📍 India