GuppyCubed FAQ overview
These answers match the FAQ on the GuppyCubed homepage.
What is a digital employee, and how is it different from an AI assistant?
An AI assistant answers one question at a time. A digital employee has a defined job, knowledge, tools and permission scope; it remembers your team's context, takes over a whole piece of work, and hands it back to a person when it is unsure.
I already use ChatGPT. Why GuppyCubed?
ChatGPT is great at one-off tasks. A digital employee remembers your team's work, continues where it left off, works with your teammates, and backs every important claim with evidence. You can also connect ChatGPT or Claude through MCP as another worker in your workspace.
What happens when an AI employee gets it wrong? Who is responsible?
It stops and asks when it is unsure, attaches evidence to what it says, and waits for your confirmation before high-risk actions such as publishing. When something is wrong, you can trace which source and version it used, and the fix goes to the right person. AI lays the problem out; people make the decisions.
How much does it cost?
The free plan covers up to 2 people with 5,000 credits a month. Monthly is US$24 per person per month, with every feature and 20,000 credits per person a month, pooled across your team. Annual is US$240 per person per year — 2 months free.
Which digital employees come by default?
Every workspace has GuppyCubed AI, a general digital employee. Five default assistants — Accounting Assistant, Sales Advisor, Brand Designer, Marketing Strategist and Customer Support — can be added in one click, renamed and re-briefed.
Can it help with client acceptance, reports and grant evidence?
Start with delivery evidence. Use projects, events, surveys and tasks to collect results, sources and open questions for clients to check — for example, headcounts against sign-in sheets and report figures against their sources. AI helps organize; the client or project owner reviews and accepts the work before billing under your agreement. Grant reporting is one use case.
Do you look at my data or train on it?
Your data is used only within your workspace. Cross-customer model improvements use only data with permission, with private content removed.