Choose a first opportunity. Build from there.

No. A meaningful problem, executive sponsorship, and access to the people who understand the work are enough to begin the conversation. An assessment can help turn that into a prioritized first build.

We agree that around the work. You need an internal owner who can help set priorities, provide context, and review results. Operators and engineers join the relevant working sessions so the system reflects their needs and they learn how to use it.

Adoption follows visible usefulness. We build around work your people already own, with an internal owner involved from the start, so the system shows up in real tasks rather than in a training exercise. Pairing and working sessions continue through delivery, and low use tells us to change the build rather than push harder on the rollout.

We meet people where they are. Engineers may pair on production tasks in their own codebase while operators begin by judging and improving output on a workflow they know well. Sessions are grouped so nobody has to learn alongside people working at a very different level.

Most teams are. We carry the engineering while your people learn on work they already need done, which keeps the learning inside their day rather than adding a separate training commitment. We agree the time involved up front and keep it to the sessions that genuinely need them.

Yes. We can carry the engineering while working with an internal owner, or pair closely with your developers. We choose a delivery approach that fits your capacity and how the system will be operated.

We aim to get something into real use early rather than after a long build-up. Quiet Light’s outreach engine now runs in daily production, and Visiting Media’s assessment turned a large body of customer evidence into a prioritized roadmap. The larger return compounds after that, as your team applies the same pattern to more of its work.

We agree where the work runs, which data it needs, and who can access it. We work with your technical team on scoped permissions and the hosting and AI accounts appropriate to the engagement, on business-grade platforms where your data is not used to train public models. Review requirements and operating controls are part of the build, and we can help write AI usage policy where you do not have one yet.

We define that for each workflow. For example, Quiet Light’s outreach system researches buyers and prepares messages for an operator to review and send. We test against real examples and decide which actions need approval before expanding automation.

Under our standard terms, client-specific custom work becomes yours once the fees due are paid. Reusable FOG tools and methods remain licensed for use with the deliverables, and third-party components have their own terms. We make the operating setup and handover clear in the engagement. Read the full terms.

Our standard engagement is month-to-month, with 30 days’ written notice to end it. We provide reasonable transition assistance on request and agree the practical handover for the systems, accounts, and documentation involved. Your agreement sets out the applicable details.

What we build comes with documentation, and the decisions behind it are written down as we go. We spread capability across the team rather than leaving it with one person, so a new hire can pick up a system without needing the person who first built it.

No. We are Claude certified because Anthropic is one of the strongest frontier labs, and that depth shows in how we use Claude Code, the Claude API, and the Claude Agent SDK. We also work extensively with OpenAI’s Codex, particularly 5.6 Sol and Astra, and use their image models heavily for visual work. We are expanding our use of Jev, where the typesafe model changes what you can depend on in production. We choose tools and infrastructure around the work, your environment, and how your team will use the result.

We update the shared queue, make the tradeoff visible, and agree what the next available capacity should address. Changes to the level of monthly investment are agreed together.