Portfolio-wide search

Effective automation depends on instant information discovery. On etg24, a single unified query searches across documents, contacts, communication histories, property notes, and the user manual — ranked by hybrid relevance in under a second.

Key search patterns

• Organization-wide discovery: An agent querying “find all records related to unit 12 at Schillerstraße” retrieves lease documents, tenant contact cards, and open maintenance tickets in a single pass.
• Strictly scoped access: The same search executed with a capability-scoped API key respects authorization boundaries automatically. A key scoped to a specific property portfolio only returns records from that portfolio.
• Hybrid search architecture: Combines lexical BM25 keyword matching with dense vector embeddings. Exact terms rank immediately, while semantic similarity catches conceptual matches and synonyms.

Architectural components

• meta.tools.search — Resolves optimal query tools for complex natural-language requests.
• meta.docs.search — Executes semantic retrieval across archived documents, bylaws, and system documentation.
• REST search endpoints: Structured filtering over properties, units, and contacts. See the Property Management v2 reference.
• Scoped API credentials: Enforce the principle of least privilege as defined in Permission scopes.

Foundational layer for AI workflows

Search serves as the underlying backbone for all other use cases: onboarding audits, inquiry triage, invoice extraction, and portfolio reporting all begin by finding the right data. Low-latency responses ensure AI agents can reason through multi-step tasks without waiting on slow queries.