The work you can hand over.
Each service is a part of your company that AI agents run for you, on a system set up for your company, with your people approving what matters. Switch on one, see how it goes, and add the next when you trust it.
Cloud and operations
The daily work of keeping systems running, and what it costs.
This is where most companies start, because it is where the cost and the risk sit. The agents do the duties a DevOps or SRE engineer would: they read your cloud accounts, keep watch, and propose changes before they make them. You get a report every week and an answer whenever you ask.
Deployments and environments
New environments, releases and rollbacks, built from code so the same thing can be made again on demand.
Monitoring and alerts
Health, quotas and spend watched around the clock, with a warning to the person who can act before something runs out.
Certificates, keys and patches
Found, listed by what expires first, and renewed or rotated on a schedule, each change shown before it is applied.
Cost
What each part of the company costs to run, where the month is heading, and what can be switched off. The provider's own figure always comes first.
Access and sign-on
One login for everything your people and your agents touch.
Most incidents are not clever. They are a shared key, a leaver who still has access, or a tool nobody put behind a login. We put one sign-in in front of your cloud consoles, servers, databases and tools, and the agents go through the same door as your people, so every session is recorded the same way.
One identity, every system
Joiners get what their role needs. Leavers lose it the same hour.
Access for a set time
Give a contractor a week on one server, or yourself an hour on a database. It is approved, and it ends on its own.
Every session recorded
Who ran what, and when, has an answer that does not depend on anyone's memory.
Compliance evidence
Ready for the assessor all year, not the week before.
The evidence an assessor asks for is mostly a record of work your company already does. When agents do that work, the record writes itself. Each finding names the requirement it sits under and the change that closes it.
PCI DSS and ISO/IEC 27001
Gaps found, a plan to close them, and policies written for how your company actually runs.
Evidence as the work happens
Access reviews, key ages, open ports, encryption, patch levels: read from your systems, not from a spreadsheet somebody keeps.
One folder for the assessor
Everything they will ask for, already collected, each item pointing to the requirement it satisfies.
AI set up for your company
The right models for each job, at a cost you can see.
Paying for one large model for everything is the most expensive way to use AI. We set up the models your company uses, from the providers you can reach in your region, and send each request to the smallest one that does the job well. Your own models can be added, and then only they are used.
Models by job
A small model for routine work, a larger one where it is needed, and a fallback that is used rarely.
Cloud and provider choice
Region, where your data may live, the evidence each provider can give you, and what the bill looks like at twice today's volume.
Cost per request
What every request cost and which model answered it, beside the provider's own invoice.
Your idea, built
For founders who know what they want and not how to code it.
Describe the business you want to build, in your own words. We shape it with you, build it, and put it on a test address you can try on your phone. Before anything goes live you get the cost of running it, on our cloud, a cloud you choose, or your own servers. Once it is live, the services above run it.
Shaped in conversation
You talk, the agents draft. You decide what stays.
A test address first
Something real to try and show people, before you commit to anything.
A clear cost to go live
What it costs to run each month, stated before you say yes.
Questions we get asked first
Do I need to know code, or anything technical?
No. You say what you want in plain words. If someone on your team does know the systems, they can step in, check a result or run a command by hand at any stage.
Can an agent change production on its own?
No. Reading runs freely. Anything that changes production, deletes something or spends money shows what it is about to do and waits for a named person. That step cannot be set to approve itself.
Does our data leave our cloud?
The part that holds your credentials runs inside your own account and sends back results, never the key. Card data and personal data are kept out of AI prompts and logs.
Do we have to take every service?
No. Each service stands on its own and can be switched on or off without touching the others. Most companies start with one.
Do you resell a cloud or an AI provider?
No. We take no commission from any provider, so when we recommend one, it is because it fits your company.
How is it priced?
Each service is quoted on its own, from what you want to hand over, and agreed before anything starts.
Not sure where to start?
Tell us what takes up your team's time. We will tell you which service would help first.