What we collect before setup
- A description of the current workflow.
- Examples of useful and unsuitable results.
- The necessary sources and applications.
- An owner who reviews questions and exceptions.
AI Dots
Implementing an AI Dot: from assignment to practical trial
A useful AI Dot starts with a concrete assignment. Learn how BigBase defines the task, prepares the setup and evaluates a practical trial with your team.
Step 1: define one assignment
“Help our administration” is too broad to test well. “Prepare a list every Monday of cases with missing attachments” identifies a source, schedule and result. Together, we make the assignment small enough to evaluate.
- Name the trigger or agreed starting point.
- Define which information belongs to the task.
- Describe the format of the result.
- Choose an employee who remains responsible.
Step 2: prepare sources and access
We identify the documents, working instructions and apps involved. We also check whether the intended setup is possible in your available workspace. Missing permissions or an unclear source need to be resolved first; the Dot should not have to guess.
A short instruction explains relevant terms, the sequence of work and how to report missing information. Examples from your business make expectations concrete.
Step 3: test the exceptions too
A useful trial includes more than an easy case:
- A normal example with complete information.
- An example with a missing attachment.
- An example with conflicting information.
- An assignment outside the agreed responsibility.
Your owner evaluates the outcome against agreed criteria. We adjust the setup within the defined scope and repeat relevant checks where necessary.
Step 4: hand over the workflow
Before everyday use, we agree who reviews results, who reports changes and what happens when an app or process changes. Handover includes the agreed instructions and practical guidance for your team.
Read about access and review and maintenance after setup. Compare the two service options on the AI Dots page.
What a completed setup provides
- A documented assignment with clear boundaries.
- A tested example of the intended result.
- Working agreements for review and escalation.
- Guidance and documentation for the responsible employee.
Frequently asked questions
How long does implementation take?
It depends on access, available examples and task complexity. We plan the steps after the intake. A date is agreed once the required workspace and sources are available.
Must the first task be fully automatic?
No. We can begin with a prepared result that an employee reviews. After a successful practical trial, we decide which further steps may run independently within the assignment.
Explore AI Dots
Implementing an AI Dot: from assignment to practical trial
Scope, preparation, trial assignments and handover for your first AI Dot.
Read this guideWhich workflows suit an AI Dot?
Examples for case preparation, research, internal work queues and meetings.
Read this guideConnecting AI Dots to apps and business information
Choose sources, avoid duplicate processing and identify the connections you need.
Read this guideAccess and human review for AI Dots
Use focused access, clear approval points and an owner for exceptions.
Read this guideAI Dot pricing: monthly service or one-time implementation
Understand both BigBase prices and how a Dot differs from an AI Employee.
Read this guideManaging and improving AI Dots after setup
Review quality, time savings and changes without quietly widening the task.
Read this guideDiscuss your first AI task
Tell us which work takes time. We help you choose a service and define the scope in a proposal.