Trimble Agent Studio
Trimble Agent Studio is an open, no-code AI agent builder available for creating and governing AI agents with deep domain expertise — no AI engineering required.
See what you can build with Trimble Agent Studio
Models, knowledge, connectors, sub-agents, evaluations and analytics — together in one no-code workspace.
Continuity: Ask your agent about a SketchUp clash today, then resume the same chat in Trimble Connect tomorrow.
Ecosystem: The agents you build in Agent Studio are instantly available in Trimble Assist and integrated products.
Interface: Built on MCP and AG-UI, Trimble Assist powers visual tools like clash dashboards and RFIs.
Put agents to work with Trimble Assist
Put agents to work with Trimble Assist
Trimble Assist is Trimble's conversational app for working with agents across the portfolio — in the Assist web app and inside integrated Trimble products.
Certifications and standards
Trimble's approach to managing AI risk is anchored on recognized frameworks, standards and regulations. The platform is certified to ISO 27001 and ISO 42001, and has passed its SOC 2 Type 1 audit, with the attestation expected in November 2026.
ISO 27001
Information Security Management System (ISMS)
ISO 42001
Artificial Intelligence Management System (AIMS)
SOC 2
Systems and Organization Controls
NIST AI RMF
AI Risk Management Framework
EU AI Act
European Union regulation of AI systems
Trimble Agent Studio frequently asked questions
Trimble Agent Studio is a no-code AI agent builder in which Trimble, its partners and its customers can safely build, evaluate, deploy, use and monitor AI agents.
What anchors Trimble's approach to AI risk management?
Our approach is anchored by frameworks, standards and regulations set outside Trimble, including:
NIST AI RMF: AI Risk Management Framework
ISO 27001: Information Security Management System
SOC 2: Systems and Organization Controls
ISO 42001: Artificial Intelligence Management System
EU AI Act: European Union regulation of AI systems
Trimble has passed independent third-party audits for ISO 27001, ISO 42001, and SOC 2 Type 1. Official certificates and the SOC 2 attestation are expected in November 2026. External penetration tests run on a regular schedule.
How are agents monitored in production?
The platform records every agent run in a standard format, so each one can be traced end to end. In Agent Studio, the Operate dashboards show usage, token consumption, errors, content filter violations, and user feedback.
What logs and audit trails does the platform keep?
The platform logs each run's inputs, reasoning steps, tool calls, and final output, so every action can be traced for forensic analysis and compliance checks.
Is customer data used to train third-party models?
No. The platform gives access to a curated set of enterprise-licensed models, served from Trimble's cloud environment. Neither the companies that build the models (such as OpenAI, Anthropic, or Google) nor the cloud providers that serve them may access, keep, or use customer data, including prompts and responses, for any purpose, including training their own models. Trimble enforces this contractually with every model and inference vendor.
Can agents use proprietary data without storing it on the platform?
Yes. Instead of uploading documents to a knowledge library, you can expose your own knowledge base or system as an MCP connector that you host yourself. The agent queries it at run time, and your data stays in your system.
How is data secured at rest and in transit?
Data is encrypted at rest with cloud-managed AES-256 by default, and every endpoint and API call requires TLS 1.2 or higher. Tenants are isolated logically, so data never crosses tenant boundaries.
Does the platform support GDPR and European data transfers?
Yes. Trimble supports compliance with the General Data Protection Regulation (GDPR) and signs a data processing addendum when a customer asks for one in writing. For personal data transferred from the European Economic Area (EEA), the United Kingdom, or Switzerland to other countries, Trimble relies on the EU Standard Contractual Clauses and other appropriate mechanisms. The Trimble Privacy Center has more information.
How do you prevent runaway agent behavior?
Agents have bounded capabilities and bounded resource use. The platform applies rate, token, and turn limits, which cap cost and protect the backend services from runaway loops and denial-of-service attacks. Agent creators can set their own usage limits on top.
How are agent integrations and tool use secured?
Calls to MCP connectors go through the Tools Service gateway, which checks the caller's identity centrally, applies least privilege, and logs each call for audit. Built-in tools (such as knowledge search and web search) run inside the platform, and local tools run in the host application that embeds Assist, under that application's own controls.
How are agents evaluated before release?
Agent Studio has a built-in evaluations module. Agent creators are responsible for running evaluations and for deciding whether their agent is good enough for its users.