FAQ

Straight answers, no sales pass

Everything people ask on a first call, including the questions where the answer is no. If something you need is missing, ask us and we will add it here.

We are pre-launch and working with design partners. Where we do not have an answer yet, this page says so rather than dressing it up.

The product

What it is, what it does, and how it differs from what you already pay for.

An AI coworker is software that owns a job end to end inside your own systems. It notices work arriving, gathers the context from every system the task touches, takes the action, and reports what it did. Nobody has to prompt it, and it handles the exceptions a script would break on.

It runs a specific job you hand it: research an account and draft the follow-up, triage a support queue and pull the customer's history out of four systems, reconcile invoices and chase the exceptions, move a request from form to approval to system of record. It is built for one job at a time rather than configured as a general assistant.

Copilots assist a person who is already doing the work, and they stop when the person stops. An AI coworker owns the work and closes the loop across systems. If your CRM is still out of date on Friday despite everyone having a copilot seat, that is the difference.

No. Deployment is done for you. You walk us through how the work gets done today, we install the coworker in your cloud, connect it to your systems, and tune it against your real cases. You are not buying a builder that needs an engineer permanently assigned to it.

The systems of record you already run: CRM, ERP, helpdesk, email, chat, databases, file storage, and internal tools that expose an API or a database. Because the coworker runs inside your network, it can reach private services a hosted SaaS agent has no route to.

Deployment and security

Where it runs, what it can reach, and how you switch it off.

Inside your own AWS, Azure, or GCP account, in a VPC you control. It is not a tenant in our platform and it is not a connector pointed at your data from the outside. It runs under the same account policy as the rest of your infrastructure.

The coworker, its credentials, and its connections to your systems all stay inside your account. You choose the model provider and control the egress path out of your own network, so what may leave the perimeter is a decision your team configures rather than one we make for you.

The install is quick, because it is a deployment into an account you already have. Building the coworker around your workflow starts with one 90-minute session, then it runs beside your team until the output stops needing corrections. That parallel period is where the real work happens.

Access to provision into the cloud account, and IAM roles scoped to the systems the job touches. Your engineers can read the infrastructure code before anything is applied, and every role you issue can be revoked without involving us.

Yes, and without asking us. Revoke the roles and access ends immediately. Stop the service or tear down the deployment and it is gone, because it was running in your account the whole time.

Buying and fit

Including the question about when this is the wrong tool entirely.

Pricing is not published yet. We are onboarding design partners and shaping the commercial model with them. If you want a straight answer for your situation, ask on a call and we will give you one rather than a range.

Operations, support, sales, and finance teams at companies of roughly 200 to 2,000 people that run a real cloud footprint and lose hours every day moving information between systems. The buying trigger is usually headcount cost; the deployment model is what makes the purchase possible at all.

Not yet publicly. We are pre-launch and working with design partners, and we would rather say that than show logos we have not earned. What we can do is walk you and your reviewer through the architecture in detail.

No. We hold no completed certifications and we do not display badges we have not earned. The security argument here is structural rather than procedural: the coworker runs in your account, under your IAM, writing to your audit logs, so there is no vendor perimeter to assess.

When a critical system has no API, no database access, and no export. When a deterministic script already handles the job, since a script is cheaper and more predictable. When nobody can describe how the process actually works. When you want one assistant for the whole company, which is a copilot. And when there is no cloud account to deploy into.

Still have a question?

These three pages answer most of what is left.

DEFINITION

What is an AI coworker?

The category, and the three things it gets confused with.

Read the definition   →
SECURITY

The page to forward to IT

Written for your reviewer, including what we have not earned yet.

Read the security brief   →
COMPARE

Against the alternatives

Agent platforms, copilots, outsourcing, in-house — and where we lose.

See the comparisons   →

Still have a question?
Ask us directly.

Bring the workflow that eats the most time and we will tell you honestly whether this fits.

Get a demo