The engine
The Codex: trained on the highest-resolution signal that exists.
The Wakefully Codex is the inference model underneath the platform. It was trained on 80,000+ annotated dreams and 5M+ structured datapoints across emotional, symbolic, belief, narrative and durability signals. Dreams were the training ground because they often carry emotional and symbolic material before it's edited into a waking explanation.
The same engine now reads waking language, reflections, intake, and partner-held data — so dreams are the ceiling of depth, not the floor of admission. A late entrant would need the same corpus. The corpus is years of trust. It cannot be compressed.
→No new collection burden. The Codex reads from what guests already provide.
→No new intake infrastructure. It routes through what the property already collects.
→The moat is the corpus. Dreams are why no one can replicate it quickly.