Glossary

Sphere terms & frameworks

The proprietary terms and frameworks Sphere uses across our work — defined once, here, so they mean the same thing everywhere you read them.

Platform Reboot™: Sphere's AI-accelerated platform transformation service: modernize legacy systems, replace aging applications, or build a custom alternative to bloated SaaS — fixed price, live in weeks, powered by Precision Driven Engineering (PDE™).
Engram: Sphere's enterprise memory layer for AI. Engram gives the model a persistent memory of each user — what they work on, who they work with, and the context they've already provided — with permission-aware retrieval and an immutable audit log.
Knowledge AI: Sphere's enterprise RAG search: a self-hosted LLM connected to internal documents through pgvector semantic search, returning source citations and confidence scores while enforcing the access controls your file system already has. No data leaves your infrastructure.
Context Tax: The hidden cost enterprises pay when AI systems must repeatedly re-load, reprocess, or re-retrieve the same information because knowledge isn't persisted or structured for reuse — inflating latency, compute spend, and error rates across every query.