VOS Team Frontier

For a team of up to twenty people whose work now runs through software and AI. We run the systems we built with you and add to them every month. Every second month we test the AI inside them against your own work, so the model doing each job is still the right one for it.

What a month holds

Frontier is everything in VOS Team Scale, with an AI Model Fit bench added on a standing schedule. Nothing in it is a one-off project. Each part repeats, and each one feeds the others. The workflows we build for you come out of your leadership calls. The bench then decides which model runs inside them.

Up to twenty seats
Everyone on the team works in the same system. We watch the automations as they run and repair them when something breaks.
A leadership call every week
Your leadership and ours, once a week, on what ran last week and what to build next.
Three new workflows a month
New pieces of work taken off your team's desks, built into the systems you already have with us.
Thirty-two posts a month
Written for your business and put out across four channels.
An org review every quarter
Your team's work read against its role-fit map, the picture built from each person's own Founder Blueprint, so work drifts back to the people it suits as roles change.
The bench, every second month
Your own test, run again against the models available that month.

A year, drawn out

Each line is one part of Frontier across twelve months. It shows how often each part comes round, which is the thing that makes a year of it different from a year of separate projects.

Leadership callevery week
New workflowsthree each month
Poststhirty-two each month
AI Model Fit benchevery second month
Org reviewevery quarter
Months are counted from the day your build goes live. The bench runs every two months on dates we set with you at kickoff, and once more if a major model ships. The day of the weekly call is also agreed before the first month, so this shows the rhythm and not your dates.

Why the bench runs every second month

New models keep arriving, and the one that tested best on your work can be beaten by a cheaper one a few months later. Most teams choose a model once, on its name, and never check again. The bill creeps up, and some answers come back wrong in ways nobody spots until they cost money.

The bench is a test built from your team's real jobs, with the right answer written down beside each one. Every second month we run it against the models available that month, including the large costly ones and the small cheap ones. For each job you get a plain answer on which model should do it and what each task costs at your volume. The test belongs to you, and it stays yours if you ever leave.

What stays fixed

The role-fit map is a picture of the team. Each person sees their own Founder Blueprint result and nobody else sees it, including you. We never use it to hire, fire, set pay, promote or discipline anyone, and that is written into our terms.

Most of what we build stops and waits for a person before anything touches money or a client. At twenty seats that matters more, and it does not change because the systems get busier.

Terms

Frontier runs a build we have done with you. It is billed monthly from the day that build goes live, and you can cancel any month. The price is fixed and published, the same for everyone.