Culture
What we believe — and how it shows up in the work
These are not values we aspired to and then forgot. They are the operating principles that cause friction when we ignore them and compound when we honour them.
Evidence over vibes
Every verdict, action, and report must trace back to observable data. If an agent cannot show its work, it does not ship. Confidence scores are not decoration — they are the contract between the system and the analyst who has to stake their name on the outcome. We distrust anything we cannot inspect.
Governance is the product
Autonomy without guardrails is a liability transfer, not a product feature. Kill switches, approval thresholds, and rollback are first-class capabilities in the ManySignal platform, not afterthoughts bolted on for enterprise compliance theatre. We built them because we would not use a product that lacked them.
Own the whole pipeline
Context lost between tools is where breaches hide. We build the detection-to-response pipeline end-to-end so nothing important falls between seams. That means resisting the temptation to call ourselves a 'platform' while quietly assuming someone else handles the hard parts.
Earn trust in stages
Detections graduate from shadow mode to supervised alerting to autonomous action based on measured performance, not marketing timelines. Trust is granted on evidence and is always revocable. This applies to how we build the product and to how we operate as a team: we prove, then we progress.
Treat analysts as the product
The security analyst who uses our platform at 3 a.m. is not a secondary consideration — they are the person we are designing for. Every UI decision, every alert format, every default configuration is evaluated by asking: does this make that analyst's night better or worse? If worse, we change it.
Write it down
Oral tradition fails at scale. We write decisions down in ADRs, post-mortems, and strategy documents that future teammates can read and critique. Documentation is not overhead — it is how asynchronous teams stay aligned, how institutional knowledge survives attrition, and how we hold ourselves accountable to what we said we would do.
Where these came from
We wrote the first version of these values in a shared document the week after we left our previous jobs to start ManySignal. They were not aspirational — they were the explicit rejection of practices we had watched fail in security operations organisations over the preceding decade.
Every one of these principles has cost us something: a feature that shipped slower because we spent the time making it explainable; a customer conversation that went poorly because we told them what the evidence showed rather than what they wanted to hear; a hiring decision that took longer because the candidate was strong but not right.
We think those costs were worth it. A security platform that people trust must be built by people who trust each other to tell the truth, document the failures, and ship things that work — not things that look like they work.
These values come to life in our people
If this is how you want to work, read our story and see the open roles we are hiring for.