01
Ship early
One module can go to market while the rest of the platform is still being built.
Evon Intelligence
Pillar Two Ventures.
We build our own AI software, spin out forensics ventures, and automate the work that slows businesses down.
“Is this genuinely AI forensics for blue-team cyber, or is it something adjacent (deception, SOAR, generic detection, red-team tooling) that only partially fits?”
If a candidate only partially fits, it is not an Evon venture, no matter how commercially attractive. A company that ships eight loosely related security tools is a portfolio; one that ships focused forensics platforms is a category leader.
Architecture
Every venture ships as independently sellable pieces.
Each venture is built as sub-modules that also compose into a full platform.
02
Sell narrow
Buyers who need one modality can buy it without taking the whole platform.
03
Compose wide
Buyers who want end-to-end coverage get a unified platform built from the same pieces.
04
Reduce risk
Incremental releases beat big-bang launches, and learning compounds across modules.
Flagship one
Is it real, or was it generated?
TryVerifAI
tryverifai.com- Category
- Multi-modal AI content forensics platform
- Core question
- “Is this document, image, audio or video real, or was it forged, tampered with or AI-generated?”
- Part of
- Evon Intelligence · Evon Ventures
Four modalities. One artifact of trust.
| Module | What it detects | Who uses it on its own |
|---|---|---|
| TryVerifAI Docs | Forged, altered or synthetically generated documents | Banks, notaries, legal, government |
| TryVerifAI Image | AI-generated images, tampered photographs, splicing | News media, insurance, KYC providers |
| TryVerifAI Audio | Voice clones, spliced audio, synthetic speech | Banks, call centres, law enforcement |
| TryVerifAI Video | Deepfake video, frame-level tampering, temporal manipulation | Media, video KYC, courts |
Each module can be bought on its own, or combined into the full multi-modal TryVerifAI platform.
Flagship two
Detection is not evidence.
Attribyte
Evidence layer for AI-agent attacks
- Category
- Evidence layer for AI-agent cyberattacks
- Core question
- “Can you prove this attack to someone who wasn’t there, and has it hit anyone else?”
- Position
- Deliberately downstream of detection, so detection vendors become partners rather than competitors.
- Built for
- Incident-response firms, insurers and forensic teams who need admissible evidence.
Certify first. Registry & Connect follow.
| Module | What it does | Who it’s for |
|---|---|---|
| Attribyte Certify | A sealed attribution certificate for each incident, built to stand up as evidence | IR firms, insurers, forensic firms |
| Attribyte Connect | Ingestion adapters into detection tools | Detection vendors |
| Attribyte Registry | A neutral registry that matches attacks across organisations | Insurers, CERTs, ISACs, regulators |
| Attribyte Seal | Cryptographic chain-of-custody infrastructure | Built into Certify |
Working on AI fraud, forensics or incident response?
We build our own AI software, spin out forensics ventures, and automate the work that slows businesses down.