Twelve questions, answered plainly

• Is my business data safe? What does the AI actually see, and where does it go?

  Only what a workflow needs. An automation that reads quote requests sees the quote request — not your whole inbox, not your accounting system — and every data path is written down in the handover. Anything we host runs in your own cloud accounts, in US data-center regions by default. For the AI itself we use established model providers, such as OpenAI and Anthropic, through their business APIs, whose terms state that data sent through the API is not used to train their models. We never use one client's data for another, and we ask for the least access that gets the job done.

  Read the security page

• Will this replace my staff?

  It removes tasks, not people. What we automate is the re-typing, the chasing, the sorting — the parts of a job nobody was hired for. The judgment stays with your team: wherever an error would cost money or trust, the automation drafts and a person approves. If a project only makes sense as a way to cut a role next month, we'll say we're not the right fit.

• What does it really cost, including ongoing fees?

  Package prices are published: the audit is $750–$1,500 one-time, automation builds are $1,200–$5,000+, websites are $1,800–$5,500 depending on tier, and the retainer is $500–$1,500 a month. Beyond that there are third-party costs we don't mark up and you pay directly: AI model usage, billed per use by the provider; hosting; and any tools you already subscribe to. Every proposal estimates them before you sign. There are no setup fees, license fees, or per-seat charges from us.

  See all prices

• What exactly do I get from the AI Workflow Audit?

  A current-state map of the workflows we reviewed; a ranked list of every repetitive task scored by payoff, effort, and risk; a plain-English call on each — automate now, later, or leave alone; fixed-price proposals for the top items; and a readout with you. The report is written so any vendor or in-house developer could act on it.

  See an illustrative excerpt

• How long does it take, and how much of my time will it need?

  The audit takes two weeks and needs about two hours a week from one point person, plus short interviews with the people who do the work. Builds take two to six weeks; you attend a short weekly demo and review results during the final parallel-run week. Websites take three to six weeks, with one review round per stage. Nobody on your side needs to write code or learn a new tool.

• What happens if the automation breaks or the AI gets something wrong?

  Three layers. Design: wherever an error would cost money or trust, the automation drafts and a person approves — nothing goes out on its own. Monitoring: every build alerts a person when a step fails, and the runbook says what to do; the old manual path still works. Support: every build includes a written post-launch fix window for anything within scope, and the month-to-month retainer covers monitoring, fixes, and changes after that.

• Who owns the code, prompts, accounts, and data? What if we stop working together?

  You do, on full payment. Everything is built in accounts in your name — cloud, database, model API keys, integrations — and the code lives in a repository you control. If we stop working together, you get a handover document and we remove our access and delete our copies of your data. There is nothing proprietary to migrate off and no ongoing license from us.

• Will it work with the tools we already use?

  Usually. We work with Google Workspace, Microsoft 365, QuickBooks, HubSpot, Stripe and Square, Calendly, and the systems common in practices, property businesses, and distributors — Clio, AppFolio, Buildium, NetSuite, Fishbowl, and others. Anything with an API or a reliable export can be connected. If a tool can't be, the audit says so before you spend anything on a build.

• You're a one-person company. What if you're unavailable?

  GainSolutions is one engineer, and everything is designed with that in mind. Builds live in your accounts with documentation a competent developer can pick up — that is the point of the ownership rule. Automations fail safe: if a step stops, a person is alerted and the manual path still works. Retainer clients get response-time commitments in writing, sized to the plan. And you are never locked in: if the founder were unavailable, you would have the code, the keys, and the runbook.

• Do I need to be technical?

  No. You need to know how your business runs and be able to give one to two hours a week during an audit or build. We explain everything in plain English, show working software rather than diagrams, and write the runbook for the person who does the job, not for a developer.

• How do I know AI will actually help my business and isn't just hype?

  You don't, until someone looks — which is what the audit is for, and it is designed to say no. We score each task on how often it happens, how clear the rules are, what a mistake would cost, and whether the data is actually available. Many tasks fail that test. Some that pass are better served by a plain automation with no AI in it, and we build that and call it what it is. If nothing clears the bar, the report says so, and you've spent a fixed, known amount to find out.

• Is there a guarantee?

  We don't advertise a results guarantee. What we commit to in writing on every engagement: a fixed price, a written scope, and month-to-month terms with no long-term contract. If the scope needs to change, it is agreed in writing before work continues.

Related: Security and data handling · Pricing · How we work