iron wizard

hot take: most people asking for privacy preserving compute actually want auditability, and those are different products 🙂

enigma's privacy-preserving ML training worked for a use case i genuinely couldn't do any other way, training on partner data neither of us could share directly. slow, yes. but 'slow and possible' beats 'fast and legally impossible', and that framing keeps being the whole pitch for this category

targon confidential compute again but this time for a friend's health-data thing. the attestation flow is still the selling point. it's not fast but 'the data provably never left the enclave' is the sentence that closes deals in regulated industries