Inquiries are open now. Paid work and onsite demos aren't open for booking yet. Prices on this site are proposed until they're confirmed in writing.
Field guide / 02
Which AI tool can your team actually use and manage?
Bring one real task and an awkward example where the facts changed. Look at the people, instructions, information, and software already involved. Then compare tools against what the work actually needs.
- What is the repeatable task? Bring three recent examples, including one awkward case.
- What is already working? Check forms, saved replies, customer records, or job-system features first.
- What information can the tool use? Decide what it may and may not see, who controls that information, and how it stays current.
- Who checks the output, and how will you test it? Name the person who can catch a bad answer before it reaches a customer or a decision. Try the tool on normal examples and the awkward one before real use.
- Who maintains it, and what happens if it fails? Account charges, access, updates, export, and exit need an owner. Agree how the work carries on when the tool is down or gets something wrong.
A practical first step: write down one task, the information it needs and who can approve the result. Bring that page to any tool conversation, with us or anyone else.
These questions overlap with the U.S. National Institute of Standards and Technology's AI Risk Management Framework (AI RMF 1.0, January 2023), voluntary guidance that NIST says is being revised. The AI RMF Core describes four related functions, Govern, Map, Measure and Manage, and says they aren't a checklist. Answering these questions doesn't make a tool, a business or Vibz Studio certified or compliant.
Tool families
Match the tool to the work.
Your existing system
If your current forms, customer-management system, documents, or job software already solve the problem, use them. Add AI only when it helps enough to justify maintaining it.
ChatGPT
It may help with drafting and repeatable tasks. Before using it for business information, check the account plan, connected apps, who can see the work, and how data is used.
Claude
It may help with long documents and work that needs shared context. Check which outside apps it can connect to and how long information is kept under your plan.
Gemini
It may fit a team already using Google Workspace. Your account settings and file permissions affect what it can reach.
Local AI
It may fit when you need a tool to work offline or have a clear reason to run it on your own equipment. Check that the chosen app is not sending work to a cloud service. Read the local AI hardware and privacy guide
I focus on one recurring business task at a time. Customer inquiries and quote follow-up are examples; I can consider another operational task after a fit review. We check the process and current software first, test AI only if it improves a defined step, and practice the result with the people who will use it. For home projects, we record how things work now, make one change, test it on your devices, and show an adult household member how to manage it.
We check current vendor documentation and the customer’s actual account before recommending access or promising a feature. A tool comparison changes when plans, permissions, or policies change.
Official documentation: ChatGPT Business data use; Claude connectors and permissions; Claude Team and Enterprise data retention; Gemini in Workspace.
If several employees need a shared AI workspace, the choice also includes account ownership, roles, connected data, spend, training and what your IT administrator must approve. See the team and enterprise AI path
See the consulting process