Your copilot can act. It can also make mistakes faster.

Build agentic systems and copilots with proper permissions, approval flows, and safety boundaries. Agents that can act without going rogue.

Agents that can read, write, and act are powerful, until they act on the wrong thing. Without proper boundaries, you get fast mistakes and slow explanations.

We design tool-using agent architectures with permissions and access control. Review and approval workflows for risky actions. Clear failure paths and containment strategies when things go wrong.

You get agents that augment human capability without creating a liability.

Tool-using agent architecture

Design agents that can use tools safely, with explicit boundaries and visibility.

Permissions and access control

Stop agents doing anything they should not. Granular permissions and enforced access control.

Review and approval workflows

Approval gates for actions that carry risk. Humans stay in the loop where they should be.

Failure paths and containment

Defined failure modes and containment strategies so mistakes do not cascade.

Ready to discuss your needs?

Book a 30-minute call

Services offered

Things That is a product engineering practice focused on building AI systems that help people make sense of complexity. For over 20 years, we've worked with teams at Google, IBM, Air New Zealand, Kpler, EE, News UK, Tesco, and The Economist, sitting in the space between product strategy and hands-on engineering. When specialist help is needed, we work with a network of senior consultants and product designers.

We're not consultants who hand off to developers. We're product engineers and designers who think strategically about what users need, then build it, from architecture to APIs to interfaces to production deployment. Our work is hands-on, writing code, reviewing pull requests, designing schemas, and testing edge cases, but always with the human experience in mind.