Intelligence

Intelligence

The group (URL /workspace/inteligencia, always visible) has a single item, AI agents. The savings meter and the soft budgets belong to the Costs group and are documented on the Observability page, under Costs.

AI agents (/workspace/agentes)

Live detection: client fingerprint, tool loop and cadence. Every piece of software talking to the gateway identifies itself: the client fingerprint, the tools the model receives and the execution loop show up here in real time, with no manual registration. Data comes from GET /api/atlasberg/agentes, refreshed every 10 seconds while the tab is visible.

• KPIs: agents active now, in agentic loop, observed actors, accumulated agent cost.
• Timeline per profile: agentic loop, with tools, conversational.
• Client type pie: SDK, CLI, framework, HTTP library, browser, unknown.
• Rankings: systems by volume and tools most used by agents.
• Agent table: family and version, profile, cadence (machine, human or undetermined), tools used and declared, requests (and not served), cost, activity sparkline, last seen.
• Expanded detail per agent: evidence (declares tools, invokes tools, returns tool results, long conversation, streaming), first seen, runtime and architecture, virtual key digest, declared app, tools with usage, models and token consumption.
• Nothing to configure. Passive observation screen.

About Atlasberg Platform

Atlasberg Platform is the control layer between a company and every AI model: each request is authenticated with a virtual key, filtered by policy and DLP, routed to the right provider and written to a hash-chained audit trail. It exposes an OpenAI-compatible API, so applications only swap the base URL.

The same artifact runs in Atlasberg Cloud, in your VPC, on-premises or fully air-gapped, and is priced by capacity and modules, never per seat.

This page is part of the official documentation. To talk to the engineering team, write to [email protected] or use https://atlasberg.com/contato. Answers come within one business day.