Skip to main content
SevrelDocumentation
← Back to App
Documentation menu

The AI Tiers

Sevrel uses an enterprise AI provider with tiered model routing. Each query is automatically routed to the right model based on complexity — no manual switching needed.

Tiered Model Routing

Commercial real estate work involves a range of document tasks — from quick tenant lookups to complex multi-property financial comparisons. Sevrel automatically selects the right AI model for each query:

  • Fast tier — quick tasks like titles, classifications, and simple lookups
  • Standard tier — standard analysis, RAG queries, agents, and general chat
  • Deep tier — deep reasoning for underwriting, complex legal analysis, and research

Model Tiers

TierBest ForCharacteristics
FastConversation titles, intent classification, memory extractionFastest responses, lowest cost
StandardDocument Q&A, multi-document comparisons, agent workflows, general conversationBest balance of speed and reasoning depth
DeepDeal underwriting, complex financial modeling, deep legal analysis, systematic researchDeepest reasoning, most thorough analysis

How Routing Works

When you send a message, Sevrel classifies the query complexity and routes it to the appropriate model tier automatically. Simple factual lookups use the faster tier, while complex analytical tasks are escalated to the standard or deep tier for deeper reasoning.

This happens transparently — you always get the best model for each specific query without needing to configure anything.

Data Security

All AI processing uses our enterprise AI provider. Data is transmitted via encrypted HTTPS connections. Your documents are read on-demand from Egnyte and are not stored by the AI provider.

Next Steps

Last updated: March 22, 2026