Services

Five ways we help you build for the long run

This page is for teams planning an AI agent, a data clean-up, a CRM change or new system connections, or looking for engineers to join them. Each service follows the same method and leaves you with something your own team can run. Together, they build a foundation your business can still stand on in 2050.

Every engagement starts with a conversation and ends with a written, fixed-price proposal. It sets out the scope, what you'll receive and the price, and you'll have it before any work starts.

Agentic AI

We build agents that act on real business data, with written guardrails, testing before launch, evaluation after, and a person in the loop where it matters.

Who it's for
Teams planning their first agent for staff or customers, and teams with an agent already live who'd like to know where it could go wrong.
What you get
  • Written rules for each agent: what it may say and do, what data it can touch, and who sees what.
  • A test suite built from those rules, including attempts to talk the agent out of them, that runs on every change.
  • The agent itself, built on the platform you already use and connected to the data it needs.
  • Evaluation after launch, with a set of realistic questions the agent is scored against, so you can see if its answers slip.
  • Clear points where a person reviews, approves or takes over, placed where the stakes are highest.
How it starts

Tell us what the agent is for and who it'll talk to. We'll reply with questions, then send a written, fixed-price proposal.

If you'd like to start on your own, our guardrails checklist is free to use.

Data

We get your data clean, connected, governed and ready for AI, on the platforms you already use.

Who it's for
Businesses whose records live in several systems that don't agree, and teams who want agents to act on data they can trust.
What you get
  • A map of where your important data lives, who looks after it and where the copies disagree.
  • Clean-up and matching rules that are written down and repeatable, so duplicates stay gone.
  • Governance you can see: who may read, change or share each kind of data, enforced in code.
  • Data shaped for AI, with the context an agent needs and nothing it shouldn't see.
How it starts

Send us a short description of your main systems and what's going wrong. We'll suggest where to start and send a written, fixed-price proposal.

Our guide, Foundations for the agentic era, covers what to get ready first.

CRM

We design and run the systems that hold your customer relationships, so people and agents work from the same picture.

Who it's for
Teams choosing a CRM, outgrowing the one they have, or adding agents to one they already run.
What you get
  • A design for your customer records, service processes and permissions that fits how your teams really work.
  • The build, on the CRM you already use or the one that suits you best, with tests for the rules that matter.
  • Screens that work with a keyboard, a screen reader and zoom, checked by tools and by hand.
  • Help with running it day to day, if you'd like it, with written notes your team keeps.
How it starts
Tell us about your teams, your customers and the system you use today. We'll reply with questions and a written, fixed-price proposal.

Integration

We connect your agents and systems through APIs, events and open protocols, including the Model Context Protocol, so they talk to each other safely.

Who it's for
Businesses with systems that need to share data, and teams connecting AI assistants to the tools their people already use.
What you get
  • Integrations built on APIs and events, with a clear contract for what each system sends and receives.
  • Connections between AI assistants and your systems over open protocols such as the Model Context Protocol, with firm limits on what each assistant can read and do.
  • Logging and alerts, so you know when a connection fails and what it affected.
  • Open standards wherever they fit, so you can replace one system without rebuilding the rest.
How it starts
List the systems involved and what should move between them. We'll reply with questions and a written, fixed-price proposal.

Forward deployed engineering

Our engineers join your team, work inside your systems and ship alongside your people.

Who it's for
Teams with a clear goal and more work than people, who'd like engineers inside the team rather than advice from outside it.
What you get
  • Engineers who work in your tools, your code and your meetings, as part of your team.
  • Work shipped in small, tested releases, early and often.
  • A planned handover, so your people own and run everything that was built.
How it starts

Tell us about the goal, your team and your systems. We'll talk it through, then send a written, fixed-price proposal.

How a forward deployed engagement runs is on its own page.

Looking for our software? Recordist and its open-source gateway are on the products page.

Where to start

Tell us what you're working on and who it's for, and we'll suggest a place to begin.