Less mystery.
Better questions.

I look at the work your business already does, find the parts where AI could genuinely help, then design and install a practical system around that. Depending on the problem, that might mean a reusable AI assistant, a structured prompt system, workflow automation, an internal knowledge tool, a website feature, training, or a combination of several pieces.

Yes. The free 30 minute consult is for understanding the business, showing you what is possible now, and identifying the smallest useful place to begin. Paid work only starts after a scope and price have been agreed.

No. You can explain the task exactly as you would explain it to another person. You do not need technical vocabulary, a process map, an API, or a preferred AI platform.

No. Choosing the tool is part of the job. The useful question is what you want to make easier, faster, clearer or more reliable. The technology should follow the problem rather than the other way around.

Bring a repeated task, slow hand-off, messy information flow, writing burden, follow-up problem, customer-service bottleneck, or process that depends too heavily on memory. One real example from an ordinary week is usually enough to start.

Depending on scope, it can include process mapping, tool selection, prompts, reusable assistants, templates, connected automations, structured knowledge, quality checks, hand-over documentation, staff training, or a website and customer-facing interface.

No. Sometimes the best answer is a clearer process, a better template or one well-built reusable assistant. I would rather install the smallest thing that solves the problem than make a simple job technically impressive for no reason.

Often, yes. Existing software, available integrations, permissions, data quality and account limits are checked before recommending a build. A new subscription is not automatically required.

No. A basic install can live inside one tool and still save a meaningful amount of friction. Connections make sense when information genuinely needs to move between systems or when repeated manual hand-offs are the problem.

Yes. A sensible first build should leave room to grow. If it proves useful, more context, integrations, automations or business knowledge can be added later rather than committing to a large system before you know what you actually value.

Not automatically. Whether personal, confidential or sensitive information is appropriate depends on the task, the platform, the account settings, the information involved and the obligations that apply to your business. Data should be minimised and the handling of it understood before it is introduced into a system.

That depends on the third-party product and the account or privacy settings being used. It should never be assumed either way. Where data handling matters, the relevant provider terms and settings need to be checked before the system is approved for that information.

Not by default. Access should be limited to what is reasonably required to build, test or support the agreed system. Client-owned accounts and the least access necessary are preferred where practical, and unnecessary access should be removed after hand-over.

No technology provider can sensibly promise that. Security depends on the services being used, configuration, passwords, permissions, integrations, staff behaviour and the sensitivity of the information involved. Higher-risk uses deserve stronger controls and more careful testing.

Sometimes, yes. The appropriate level of disclosure depends on the use. A customer-facing chatbot should be clearly identifiable as AI, while an internal drafting tool may not need the same disclosure. Transparency matters more as the system has more influence over a person or their information.

Not automatically. AI can sound certain while being incomplete, outdated or wrong. Important facts, calculations, commitments and external communications should have a level of checking that matches the consequences of getting them wrong.

Work dominated by sensitive judgment, legal or safety consequences, unclear responsibility, unreliable information or decisions that materially affect people may be unsuitable for full automation. Sometimes the correct implementation is to assist a human rather than replace the decision.

Some routine, low-risk tasks can operate with very little intervention. Higher-impact work should keep suitable review, approval or override points. The goal is not maximum automation; it is the right amount of automation for the consequence of the task.

Your business remains responsible for how it uses the tools and for decisions made from their output. An AI system can recommend, draft, classify or flag, but it does not take legal or managerial responsibility away from the people operating the business.

Third-party AI models and software change over time. Features, limits, behaviour, pricing or integrations can move after an installation. Important systems should be checked when a provider makes a material change, and larger setups may need occasional maintenance or refinement.

The first basic AI install is $99, additional basic installs are $150 each, and the full five-install starter pack is $650. Custom AI setups range from $200-$800, premium websites or landing pages from $300-$500, AI business audits are $150, and simple automation add-ons start from $250. Every paid engagement has an agreed scope and price before work begins.

Most likely. Some products have free plans; others charge monthly, per user or by usage. Any third-party cost that is relevant to the proposed setup should be identified before you approve the work.

For a one-off Sorrin build, the setup delivered to you is yours once the agreed work has been paid for, subject to any third-party platform licences, provider terms and rights that sit outside Sorrin. You are not required to take a membership simply to keep using a paid one-off build.

No. Memberships are optional ongoing support. They make sense when you want continued access, check-ins, refinements, strategy or more active technical involvement after the initial work, and can provide greater overall value when you are planning multiple installations.

The hand-over should make the system understandable enough for you to operate it. Depending on scope, that can include documentation, training, testing notes and agreed refinements. New work, major changes or ongoing maintenance sit outside the original scope unless they were included.

Yes. Practical one-on-one owner or staff training is available from $250/hr and is focused on the tools, decisions and techniques that apply to the work your team actually performs.