Glossary
Key concepts and terminology used in the Native AI platform.
Business Learning Model (BLM)
An AI system that continuously learns from a specific business's data — emails, documents, conversations, calendar — to build a contextual understanding unique to that business. Unlike general-purpose AI, a BLM improves with use and reflects the specific patterns, relationships, and preferences of the organization it serves.
Mind Graph
A structured representation of business entities (people, companies, projects) and their relationships, automatically built and maintained from integrated data sources. The Mind Graph updates continuously as new data arrives and serves as the foundation for all AI features in Native AI.
Multi-Step Orchestration
The ability to decompose a complex user request into a plan of sequential and parallel steps, each executed by specialized AI agents with scoped tools. For example, a request like 'Schedule a follow-up with the client we met last week and draft a recap email' is broken into contact lookup, calendar check, event creation, and email drafting — all handled automatically.
Task Extraction
Automatic identification of actionable tasks from team conversations and emails, including due dates, assignees, and priority levels. Extracted tasks are surfaced in a central panel with bulk management capabilities, ensuring nothing slips through the cracks.
Insights Engine
A system that automatically surfaces business insights from communication patterns — key decisions, action items, relationship signals, and trends — with deduplication and relevance ranking. Insights are verified and updated as new data arrives.