Privacy Policy

Introduction

RecruitBase (“Company,” “we,” “us,” “our”) builds and maintains AgentFit — an open-source AI agent evaluation and interpretability framework — and operates the website at recruitbase.work. This Privacy Policy explains what data we collect, how we use it, and your rights in relation to it.

Self-hosted framework: If you download and run AgentFit on your own infrastructure, RecruitBase does not collect or process any of your evaluation data. This policy applies to use of this website and any managed enterprise service offered by RecruitBase.

Information We Collect

Website visitors

• Anonymous analytics via Umami (self-hosted, no cookies, no personal data shared with third parties)
• Information you voluntarily submit through contact or demo request forms (name, email, company)

Managed enterprise service accounts

• Account information (name, email, company, role)
• Billing and subscription information
• Team members, roles, and permission settings

Evaluation data (managed service only)

• Business Need Profiles (BNPs) you define
• Agent evaluation scenarios and test cases
• Dimension scores, sub-metrics, and overall evaluation results
• LLM interpretation outputs and recommendations
• Evaluation audit trail events (timestamped, immutable)
• Exported evaluation reports

How We Use Your Data

• Providing, operating, and improving the RecruitBase website and managed Service
• Running agent evaluations across seven behavioral dimensions as requested
• Maintaining immutable evaluation audit trails for reproducibility and governance
• Generating and delivering evaluation reports on request
• Responding to demo requests and support inquiries
• Sending service updates and notifications (managed service accounts only)
• Anonymous, aggregate analytics to improve the website
• Complying with legal obligations and responding to lawful requests

LLM Interpretability and Third-Party Providers

AgentFit's interpretability layer sends evaluation data — scores, sub-metric breakdowns, weighted arithmetic, and BNP context — to a Large Language Model to generate explanations and recommendations. Important notes on how this works:

• Self-hosted framework: You choose and configure your own LLM provider. Data is sent directly from your infrastructure to that provider. RecruitBase is not involved.
• Managed service: We support multiple LLM providers (OpenAI, Anthropic, Google, Groq, Ollama, and others). Where a provider is used, evaluation context — scores, BNP weights, and anonymized agent metadata — may be transmitted to that provider under their data handling policies. We do not transmit your account credentials or other personal information to LLM providers.

Data transmitted to LLM providers is minimized to what is functionally necessary for the interpretation feature.

Evaluation Audit Trail

Evaluation records — BNP definitions, dimension scores, interpretation outputs, and report exports — are stored as immutable, timestamped events. This immutability is a core design property, enabling reproducibility and governance auditability. These records cannot be edited or deleted by users in the managed Service. You may export the full evaluation record at any time as JSON or PDF.

Website Analytics

This website uses Umami, a privacy-first, self-hosted analytics tool. Umami does not use cookies, does not track users across sites, and does not share data with third parties. The analytics data we collect is aggregate and anonymous — page views, referrers, and general device type. No personally identifiable information is collected through analytics.

Data Security

We implement industry-standard security measures to protect data processed through the managed Service, including encryption in transit (TLS) and at rest, role-based access controls, and scoped API permissions per team member. If you are self-hosting AgentFit, the security of your evaluation data is entirely under your control. No internet transmission is 100% secure; report any suspected breach immediately to [email protected].

Data Retention

We retain managed Service account data for as long as your account is active or as needed to provide the Service. Evaluation audit trail records may be retained beyond account deletion for legal compliance purposes. You can request deletion of your account and associated evaluation data by contacting us at [email protected], subject to applicable legal retention requirements.

Open-Source Framework and GitHub

The AgentFit framework is publicly available on GitHub. Any interactions you have with the GitHub repository — issues, pull requests, discussions — are governed by GitHub's Privacy Policy and Terms of Service. RecruitBase does not collect personal data through the GitHub repository beyond what GitHub makes available to repository maintainers (contributor usernames, issue content).

GDPR, CCPA, and Your Rights

Depending on your location, you may have the right to:

• Access the personal data we hold about you
• Rectify inaccurate or incomplete data
• Request erasure of your account data (subject to audit trail retention obligations)
• Object to or restrict certain processing activities
• Receive a copy of your data in a structured, machine-readable format
• Withdraw consent for optional processing at any time

To exercise any of these rights, contact us at [email protected].

Changes to This Policy

We may update this Privacy Policy from time to time. We will indicate the revision date at the top of this page. Continued use of the website or managed Service after changes take effect constitutes acceptance of the updated policy.

Contact

For privacy questions or to exercise your data rights:

Email: [email protected]
GitHub: github.com/RecruitBase/agentfit/issues