Managed AI, the way you already buy everything else that matters.

ScaleSight AI prices managed AI per seat plus usage. Each seat carries a monthly usage allowance, and you set the ceiling on total spend. Our team sets the allowance for each team and department at setup, reviews it against real use every month, and brings you the changes worth making. You see spend by team and by department while the month is still running.

• Per seat plus usage, said out loud. Seats are the predictable part of the bill. Usage is the part that moves, and it moves inside a ceiling you set. ScaleSight AI is not flat and it is not unlimited, and you hear that before the contract rather than on the first heavy invoice.
• You decide who gets a seat. A seat costs money, so you seat the people whose work happens on a screen. In a manufacturer or a field services business that is usually the office. ScaleSight AI recommends going a little wider than the people who asked loudest, because useful patterns often come from people who would never have asked, and then reports which seats went unused so you can drop them.
• Different teams, different allowances. Sales writing proposals all day draws more than finance closing the month. ScaleSight AI sets an allowance for each team and department at setup, and changes it when the work changes. Budget follows the work, not headcount.
• We watch the allowances, you approve the changes. ScaleSight AI flags which team is close to its allowance while the month is still running and brings you one recommendation: raise that team, move unused allowance across from another, or leave it alone. You approve. You do not have to monitor anything.
• Cheaper models for routine work. Nothing ScaleSight AI builds is tied to one AI vendor. Routine drafting runs on whichever model does it well for the least money, and the expensive models are kept for work that needs them. It is one of the levers that keeps a month’s work inside its allowance.
• One bill instead of nine. The scattered personal subscriptions come off the expense reports. What replaces them is a seat charge you can predict and a usage charge you can watch, on one invoice, under a ceiling you chose. ScaleSight AI builds the list of what can be canceled during setup.

Because the work happens in one place, ScaleSight AI can report what kinds of work are going through the workspace, by department, in plain language, and review that report with you. Our team also defines the review points: which outputs a named person checks, and what AI is not allowed to do unsupervised.

• What the work is, by department. ScaleSight AI reports the categories of work going through the workspace, ranked, so you can decide what to standardize and what to stop. It covers work inside the workspace. Anything still happening on a personal account outside it is invisible, which is the practical argument for moving it in.
• Where a person signs off. AI gets things wrong. Anything with money or a customer on the other end gets a review point: quotes, estimates, pricing, client-facing documents, anything with a number in it. ScaleSight AI writes down, per workflow, what the AI does, what the person checks, and what the AI does not do on its own.
• Reporting you can act on. A login count tells you how often somebody opened an app. It does not tell you what went through it. ScaleSight AI reports the kinds of work, which is the thing you can make a decision about.
• Rules set once, applied in the workspace. Who can use what, and which connected sources they can reach, configured once by ScaleSight AI instead of relying on everybody remembering a policy. A rule can control access and sources. It cannot read intent, so it works alongside the review points rather than replacing them.
• Security posture, stated correctly. ScaleSight AI runs on a SOC 2 Type II audited platform, hosted in AWS US regions, with each customer tenant isolated, no training on your data, and outbound calls to model providers under near zero retention settings. We send those written terms before you sign, and we say plainly which certifications are held by the underlying platform and which by ScaleSight AI as a company, because a security questionnaire will ask and the two are not the same thing.
• Something to hand your customers. When a customer contract, an insurer, or a security questionnaire asks how you use AI on their data, you answer from the workspace record and the written platform terms rather than from a policy document nobody has checked against reality.

ScaleSight AI runs one workspace the company owns instead of separate personal accounts. Shared prompts, shared projects, and published templates are visible to colleagues and reusable by them. Our team also writes down in plain language what colleagues and administrators can see.

• You choose what gets shared. What colleagues see is what a person puts in a shared space. Private work stays private to that person. ScaleSight AI writes this down during setup so you can say it out loud in a team meeting without guessing.
• It stays when the person goes. How your estimator gets AI to help price a job belongs to the company once it is published, and it stays there when the laptop walks out the door.
• New hires start where the company is. Somebody joining in month nine opens what the company has already worked out, on day one, instead of a blank box and a login.
• Departments can borrow from each other. The way service answers a difficult customer email is most of the way to how sales answers a difficult customer email. In separate personal accounts, neither one ever finds out.
• Nobody has to hide that they used AI. When AI is simply how the company works, there is nothing to defend. People who were quiet about it stop being quiet, and what they figured out gets used.
• Your people, spreading their own work. This pillar is your team spreading its own work. New templates, agents, and workflows built by ScaleSight AI’s team are a separate thing, covered under Acceleration.

ScaleSight AI’s team keeps building on the managed service. Some of what arrives is built from the jobs your own usage report ranks highest. Some of it is work that is the same in every business, built once and given to every client. Our team also re-points and re-tests what is already there as models and prices change.

• New work every month. Working templates, agents, and workflows, handed over ready to use, whether or not anybody at your company thought to ask for them.
• Shared library, private work. The library is the work that is the same in every business, built once and given to every client. Anything built on your documents, your data, or the specific way your company does a job stays in your workspace and is never shared with another client.
• Built from what your people actually do. The usage report says which kinds of work come up most often in your company. Those are the ones ScaleSight AI builds first, so we are not guessing at your priorities.
• Work that is the same in every business. Proposals, meeting notes, job postings, quotes, follow-ups, onboarding packets. ScaleSight AI builds those once and every client gets them, so you are not paying to solve a problem that has already been solved.
• We maintain what we build. Models change and prices change, and a workflow that worked in January drifts. Re-pointing it to a better or cheaper model and re-testing it against the cases it was accepted on is our job on the managed service, rather than something you have to remember to schedule.
• Training is included, and it is short. Somebody from ScaleSight AI shows your team how to use each thing we hand over, in short sessions we record so a new hire can watch them later. You do get training. You do not get a quarter-long change program.
• Where the managed service ends and a project starts. In the managed service: templates, agents, and workflows that run inside the workspace on material you upload, plus maintenance. A separate AI Build Project: anything that connects to another system such as an ERP, CRM, or ticketing tool, anything that writes back into one, and anything that needs custom testing against your data. ScaleSight AI says which one it is in writing before work starts.