Docs/Overview

ANTS Platform Documentation

One place to see, control and govern how your organization uses AI: trace your applications and agents, discover the AI tools employees use, manage cost and budgets, and produce evidence for compliance.

New here? Start with the Quick Start to create a project and API key and send your first trace. See what changed in the Release Notes and what is coming on the Roadmap.

What is ANTS Platform?

ANTS Platform is an AI governance platform built on agent observability. Your applications and agents send traces through the SDKs, OpenTelemetry or a connector. Endpoint agents and a browser extension report the AI tools people use. Everything lands in one organization, split into projects, with role-based access to every page.

Start hereGetting Started

Create a project and key, send your first trace.

LearnCore Concepts

Organizations, projects, traces, agents and credits.

Plug inIntegrations

SDKs, frameworks, connectors and alerting tools.

ReferenceAPI Reference

Public REST API and authentication.

Platform Overview

The platform is organized into modules that match the sidebar of the app. Each can be adopted on its own, and access to each is controlled by roles and custom roles.

Observability

Traces, observations, sessions and scores for every request your applications and agents make, with cost, tokens and latency. Send data with the Python, JavaScript or Java SDK, the OpenTelemetry endpoint, or a trace connector. See Observability.

quickstart.pycopy
from ants_platform import AntsPlatform
from ants_platform.openai import openai

ants = AntsPlatform()  class=class="tok-s">"tok-c"># reads ANTS_PLATFORM_PUBLIC_KEY, _SECRET_KEY and _HOST

with ants.start_as_current_span(name=class="tok-s">"support-request", agent_name=class="tok-s">"support_agent"):
    openai.chat.completions.create(
        model=class="tok-s">"gpt-4o-mini",
        messages=[{class="tok-s">"role": class="tok-s">"user", class="tok-s">"content": class="tok-s">"How do I reset my password?"}],
    )

ants.flush()

Discover AI

Find the AI tools, desktop apps, coding agents, MCP servers and AI agents in use across the organization, by user, department and device, with a registry and policy for approved tools. See Discover AI and the AI Agent Center.

Gauge AI and Budget AI

Gauge AI covers cost and latency dashboards, model prices and the AI Optimizer for comparing models on your own traces. Budget AI is the LLM gateway: provider credentials, model routing, keys for people and teams, and budgets that alert or stop requests. See FinOps and Budget AI.

LLMOps

Prompt management with versions and labels, a playground, datasets, experiments, evaluations and annotation. See LLMOps.

AI Governance

AI Guardrails: per-agent policies that log, block or redact PII, PCI, PHI, harmful content and jailbreak attempts in prompts and responses.
AI Inventory and approvals: approve the models agents may use and see every agent that used an unapproved one.
Risk and drift: a risk score per agent, and alerts when what an agent does no longer matches its declared controls.
Compliance evidence: evidence in eight categories, review through Slack, Jira or Linear, coverage of the EU AI Act, ISO/IEC 42001 and NIST AI RMF, and PDF or Excel reports.
Access and audit: built-in roles (Owner, Department Admin, Security & Compliance, Cost Analyst, Auditor), custom roles, SSO, SCIM and an audit log.

See AI Governance and Administration.