<?xml version="1.0" encoding="UTF-8"?><oembed><type>video</type><version>1.0</version><html>&lt;iframe src=&quot;https://www.loom.com/embed/0d833a2bcff943cab89c4499e93be817&quot; frameborder=&quot;0&quot; width=&quot;1016&quot; height=&quot;762&quot; webkitallowfullscreen mozallowfullscreen allowfullscreen&gt;&lt;/iframe&gt;</html><height>762</height><width>1016</width><provider_name>Loom</provider_name><provider_url>https://www.loom.com</provider_url><thumbnail_height>762</thumbnail_height><thumbnail_width>1016</thumbnail_width><thumbnail_url>https://cdn.loom.com/sessions/thumbnails/0d833a2bcff943cab89c4499e93be817-f658bb9780902b83.gif</thumbnail_url><duration>378.63466600000004</duration><title>Veriflow AI: Verify Facts, Not Documents</title><description>This Loom explains Veriflow AI, a privacy-first identity verification platform for the Flare Summer SIGNA hackathon that verifies facts instead of storing raw documents. It uses client-side encryption with AES-256-GCN via WebCrypto, performs decryption and OCR inside an attested TEE enclave, wipes PII from enclave RAM after checking the rule (for example, passport age 18+), and returns a cryptographically signed Boolean proof. The proof is anchored to deployed Veriflow registry smart contracts on Flare Coston and Coston two testnets for independent on-chain audit. A developer portal provides API keys for third-party dApps to receive signed responses tied to smart contracts on Flare using S-I-W-E wallet authentication.</description></oembed>