Offensive security, done properly

PurpleSable runs advanced red team operations and penetration tests for organisations across multiple sectors. Direct involvement of the same team across the entire length of the assessment, from scoping to the final report.

Three ways to find out how an attacker would get in

AI supports our testers, never replaces them

AI can automate parts of a penetration test, and it keeps improving. It still falls short of an experienced tester, who knows which findings matter, how to chain them into a real attack and when a result is a false alarm. So we use AI to extend our operators' skills and speed up their work, not to do the testing for them.

How we use AI

Where AI helps

  • Building custom offensive tooling and developing proofs of concept faster
  • Mapping attack paths and pinpointing high-value targets
  • Extending operator knowledge of unfamiliar technologies
  • Consistent, efficient report writing

Our commitments

  • Your data is never used to train AI models
  • Company names and identifying details are anonymised, so they never reach a model
  • AI assists our testers and never conducts the test

How an engagement runs

  1. Free 45-minute scoping call
  2. Proposal and fixed quote
  3. Signed contract and rules of engagement established
  4. Testing, with agreed check-ins and stop conditions
  5. Report and debrief
  6. Retest of fixed findings

The full process

Start with a 45-minute scoping call

No sales deck. You talk to an experienced operator, and you leave with a clear scope and a sense of whether we are a good fit.

Talk to an operator