The frontier AI for IT work
The frontier AI for IT work
You sell a flat monthly fee. Your cost of delivery is almost entirely labour. Theo resolves L1 tickets end-to-end so the next junior you were about to recruit becomes a hire you defer rather than a cost you carry.
Your technicians are the advantage. They know the clients, make the judgement calls, and do the project and escalation work that actually grows the business.
Ticket volume still drifts up with every new application a client adopts. Hiring another junior to absorb that queue keeps you in place at a higher payroll. Theo takes the repeatable L1 work — password resets, account unlocks, licence assignments, mailbox permissions — so the same people can take on more seats, more clients, and more of the work only they can do.
More tickets per technician means room to grow without adding headcount just to keep pace.
Keep experienced people on client conversations, escalations, and projects instead of resetting passwords between them.
When routine tickets are handled, technicians can deliver the thoughtful fixes and proactive guidance that deepen client relationships.
Theo works inside the stack you already run. Your team supervises it and expands autonomy without becoming workflow builders.
Take on more work with the people who already know your clients, then hire because the opportunity demands it — not because L1 volume does.
This is the decision in front of you when the queue is full. Both options add capacity. They do not cost the same, and they do not carry the same risk.
| Hire an L1 technician | Deploy Theo | |
|---|---|---|
| Annual cost | $45,000–$55,000 plus benefits, payroll tax, licences and a seat | Less than the junior you were about to recruit. Exact pricing depends on your endpoint count and ticket mix — we go through it on the call. |
| Time to productive | 4–8 weeks to hire, then a few months to competence | 1 week to live, 1 month to ROI |
| Overnight and weekend coverage | No, or a shift premium | Yes, at no additional cost |
| Attrition risk | Real, and concentrated in the first two years | None |
| Repeated ramp cost per hire | Paid again on every replacement | Paid once, never again |
| Capacity when volume spikes | Fixed at one person's throughput | Scales with the queue |
| Consistency across tickets | Varies with the day and the person | The same loop on every ticket |
| Handles judgement calls and client relationships | Yes — this is what you hire people for | No, and it escalates rather than guessing |
| Setup fee | Recruiter fee, or your own time | None |
We are not going to publish a number that turns out to be wrong for your business. Pricing depends on managed endpoints and ticket mix, and it is a five-minute conversation on the first call.
Theo isn't here to replace your technicians. They're the people clients trust, and they do the project and escalation work that actually grows the business.
What changes is the next job posting. The L1 tickets that role would have covered get resolved without it — so you can grow with the team you have, and hire when the opportunity is worth another seat.
Your technicians become agent managers. They supervise Theo the way they would supervise a junior, and they can overrule it on anything. The frame is that they gain a report, not that they lose a job.
We would rather tell you on the first call than six weeks into a deployment.
Less than the junior you were about to recruit. We price on managed endpoints and ticket mix, so a number on a website would be wrong for most people reading it.
We go through exact pricing on the first call, along with which categories you would grant autonomy on and what that removes from the queue. There is no setup fee and no services engagement.
You can — at $45,000–$55,000 plus benefits, four to eight weeks to hire, and a few months to competence.
Theo costs less, starts in a week, works the overnight queue, and doesn't leave in eighteen months. Most of our MSPs use it to defer the next hire rather than to reduce the team they have.
One week from discovery call to first resolved ticket, and a month to ROI in the usual case. The first month is largely supervised, so the return in that period is your senior technicians handing junior work off rather than a headcount change.
The clean return arrives when you decide not to post the next req.
A ticket counts as automated only if it was resolved end-to-end with zero human touch and closed by Theo. Triaged, drafted, escalated and human-approved tickets do not count.
We publish the measurement window, the number of MSPs and tickets in the sample, and the breakdown by ticket category so you can map it to your own mix rather than take an average on faith.
That is your call and your contract. What we can tell you is that the audit trail is built for the conversation: every action is timestamped, attributable and exportable, so you can answer a client's question about a specific ticket precisely rather than approximately.
Less than you expect, if you introduce it as a report rather than a replacement. Theo takes the categories nobody wants and shows exactly what it did; the technician approves, corrects or overrules.
The teams where this goes badly are the ones where autonomy was switched on without anybody watching it work first. That is why week one is fully supervised.
Your client data is not used to train models. Full stop.
SOC 2 Type II — Trust Center. Data residency, retention window and the subprocessor list are published in the trust pack we send before any pilot.
Theo acts through the same delegated permissions your technicians use, scoped per client. You can revoke access from your side at any time without going through us.
Book 30 minutes. We'll go through your endpoint count, your ticket mix and the actual price, and tell you whether deferring that hire is realistic this year.