FAQs

Infiltra.ai is an automated penetration testing platform built around specialized autonomous agents that replicate the methodology of an experienced human pentester.

No. Infiltra is a cloud based platform. Scans run on demand or on schedule from your selected region.

Web applications (black-box and authenticated) and APIs. For APIs, we recommend importing an OpenAPI (Swagger) definition file or Postman collection to map endpoints for testing.

Yes. Infiltra includes "Production Mode" guardrails that exclude disruptive exploit vectors. You can also configure request rate throttling (RPS) to ensure the scan stays within your infrastructure's load capacity.

Absolutely. You can schedule assessments to run daily, weekly, or on a custom recurring basis, or trigger them via API/Webhooks as part of your CI/CD/CS process for continuous assurance.

We implement Tenant Specific Encryption Key (TSEK) architecture. This ensures cryptographically isolated environments for all sensitive configuration metadata. In addition, each scan runs in an isolated processor environment which is destroyed on scan completion to avoid context spill between scans or customers.

SSO is on the roadmap. You can export findings as JSON/CSV today. Native Jira/ServiceNow integrations are planned.

We provide persona-based reporting that includes executive summaries for leadership and technical reproduction evidence, complete with payloads, to help developers remediate findings quickly.

OWASP Top 10 and SANS 25. Findings are severity-scored using CVSS and mapped to categories for reporting.

Standard scanners identify vulnerabilities based on signatures. Infiltra uses autonomous agents that attempt to chain and execute exploits in a safe way to provide evidence of impact. Findings can be marked as False Positive or Closed, preserving an audit trail.

While vulnerability signatures are regularly updated, our team also monitors various threat intelligence feeds to ensure our agents are regularly updated with the latest offensive security research and attack vectors.

We have specifically implemented Browser agents with vision models to handle complex authentication workflows by managing session states and persistent logins, allowing for deep, automated testing even in environments protected by Multi-Factor Authentication.

Yes. Infiltra Agents attempt horizontal and vertical privilege escalation when a foothold is established, and re-scan from the elevated context where applicable.

Yes. Our autonomous agents perform multi-step reasoning to replicate a human pentester's methodology, specifically testing for logic flaws like IDOR by attempting unauthorized context switches.