Research notes for verifiable AI.
The CORTEX blog is where product positioning becomes technical argument: why memory needs proof, how agentic systems fail, and how blockchain/NFT audit evidence can be made reviewable.
Agent memory
Read essay
Why AI memory needs a ledger, not just recall.
Vector recall helps agents remember. It does not prove what was admitted, when, by whom, or under which guard.
Crypto/NFT audit
Read essay
From wallet chaos to advisor-ready evidence.
How CORTEX can structure onchain activity into a technical package for fiscal review without pretending to be tax advice.
MOSKVbot
Read essay
The memory passport pattern for serious agents.
Agents need continuity, but continuity without provenance becomes liability. The passport pattern fixes the boundary.