Most legal AI treats every conversation as if it's the first. Calliope Strickland builds AI that remembers—not through brute-force storage, but through intelligent memory systems that capture what matters and surface it when needed.

At Caddy, Calliope led the AI team that built consumer-facing AI agents from the ground up. She designed the RAG pipelines, custom embedding models, and ML-ops tooling that made those agents actually useful—not just impressive in demos. This is exactly the architecture that powers Memoria's memory engine.

At Flipboard, Calliope built deep-learning recommendation systems that learned what users actually wanted—often before they knew themselves. This isn't just technical skill; it's understanding how to build AI that surfaces the right information at the right moment. For a law firm, that means surfacing the relevant precedent, the partner's preferred approach, the clause that won the last negotiation.

At WireWheel, Calliope designed ML models that monitored privacy compliance without exposing sensitive data. Building AI that's both intelligent and private isn't a contradiction—it's a design discipline she's practiced for years. This is why Memoria can capture institutional knowledge while maintaining zero data retention.

Memoria's 'hierarchy awareness'—the ability to surface senior partner preferences to junior associates—isn't magic. It's the result of Calliope's work on preference learning and contextual retrieval. 'Write it your way, then apply their memory' works because of the underlying ML architecture she designed.