Start with access, not the AI
Before any agent runs, write down what it will read, what it can change and who can switch it off. Give it the minimum access needed for its job and nothing more.
What good data handling looks like
- Data is encrypted.
- Access is limited to authorised team members under role-based controls.
- Customer data is never sold.
- Data can be deleted on request.
- Where the data is stored is stated clearly. Ours is stored on servers in India.
Keep a record and a human in the loop
Every call, message and follow-up should be logged so you can see exactly what the agent did. Anything unusual should route to a person rather than being guessed at.
Questions to ask any vendor
- Which of our systems will you access, and why?
- Who on your team can see our customer data?
- Where is it stored, and for how long?
- Can we see a log of what the system did?
- How do we revoke access and delete our data?
Our approach
We start with an audit of your tools and processes, agree the access the agent needs, and keep monitoring it after launch. If you are not comfortable giving access to a particular system, say so on the call and we will work around it.
Want help automating your calls and follow-ups?
Book a live demo and find out where automation would pay off fastest.
Book a live demo