Skip to main content
Stardex integrates AI across the platform to help recruiters work faster and smarter. This article explains where AI is used, which providers power it, and how your data is handled.

​
Where AI Is Used

Search & Matching Natural language candidate search, semantic candidate-to-job matching, similar profile recommendations, and job suggestions for candidates. Resume Parsing Automated extraction of structured candidate data — work history, skills, education, and contact information — from uploaded documents. Content Generation AI-drafted candidate summaries, job specifications, outreach email sequences, and personalized email merge variables. Content is streamed in real time as it’s generated. Research & Sourcing AI-suggested target companies, candidate identification at specific organizations, and auto-generated research briefs to support sourcing strategies. Smart Automation Automatic candidate tagging, email sentiment detection, AI-assisted data mapping during imports, follow-up task suggestions from meeting notes, and calendar-to-pipeline stage tracking. Candidate Chat A conversational interface where recruiters can ask natural language questions about a candidate’s profile and receive contextual answers. Meeting Intelligence Automatic transcription and summarization of recorded calls and meetings, with suggested follow-up actions.

​
AI Model Providers

Both providers are accessed server-side through their enterprise API tiers. No AI processing happens in the browser.

​
Enterprise Security & Data Privacy

No data retention by AI providers All AI requests go through the enterprise API tiers of Azure OpenAI and Anthropic Claude. Under these tiers:
  • Prompts and completions are not stored by the model providers.
  • Customer data is never used to train or improve any AI models.
  • Requests are processed and discarded — no logs of inputs or outputs are retained on the provider side.
Server-side processing only All AI calls are made from Stardex’s backend servers. No candidate data, client data, or user content is sent to AI models directly from the browser. The browser only receives the generated output. Data isolation Customer data is stored in a hosted PostgreSQL database with row-level security (RLS) enforced per organization. Each organization’s data is logically isolated at the database level, ensuring one customer’s data is never accessible to another. Encryption Data is encrypted in transit (TLS) and at rest. AI API calls to Azure and Anthropic are made over encrypted HTTPS connections. Provider compliance
  • Azure OpenAI operates under Microsoft’s enterprise compliance framework, including SOC 2 Type II, ISO 27001, and GDPR commitments. Azure OpenAI’s data privacy terms explicitly state that customer data is not used for model training.
  • Anthropic Claude (API tier) does not train on customer data and does not retain inputs or outputs. Anthropic maintains SOC 2 Type II certification.

​
Frequently Asked Questions

Does Stardex send my data to OpenAI or ChatGPT? No. Stardex uses Azure OpenAI, which is Microsoft’s enterprise deployment of OpenAI models. This is separate from the consumer ChatGPT product. Your data is not shared with OpenAI’s consumer services and is not used for model training. Can I opt out of AI features? AI features are integrated into the platform to enhance workflows, but most are triggered by explicit user actions (e.g., clicking “Generate Summary” or running an AI search). Passive AI features like auto-tagging and email sentiment detection can be discussed with your account team if you have specific requirements. Where is my data stored? Your core data is stored in Supabase (hosted PostgreSQL). AI-generated content (summaries, job specs, etc.) is saved back to the same database and subject to the same security policies. No customer data is stored with AI model providers. Is data shared across customers? No. Row-level security ensures complete data isolation between organizations. AI requests are made in the context of your organization only.