Automated follow-up calls and WhatsApp reminders, on rules you set
A patient booked a slot. A buyer asked for a callback. An applicant stopped halfway through a form. The agent calls or sends a WhatsApp message at the time you set, notes the answer and passes anything unusual to your team. This page covers reminders and callbacks for existing contacts: people who booked, enquired, applied or ordered.
Follow-ups we can build
- Appointment reminders. A call or message before the visit, with an option to confirm or reschedule. Clinics are one example (clinic bookings).
- Callback requests. Someone ticked “call me back” on a form. The agent calls, answers the basics and books time with your team.
- Unfinished applications. A reminder about missing documents or steps, with the list of what is still needed.
- Order confirmations. A check before dispatch, for example on cash-on-delivery orders (online retail).
The rules you set
Before launch, we agree:
- the hours you choose for calls and messages,
- how many attempts it makes and how far apart,
- whether it messages on WhatsApp first or calls first,
- what counts as done, so nobody gets a reminder for something already handled.
Stopping, and handing over
We can set the agent so that anyone who says they want no more calls is marked and taken off the list. If they want to talk to someone, or their question falls outside the script, the agent books a callback with your team and passes on the notes.
What you see afterwards
Each attempt gets an outcome: reached and confirmed, rescheduled, no answer, asked to stop, or needs a person. Outcomes can go to your CRM where it offers an API (integrations), or to a shared sheet.
WhatsApp reminders
We build messaging bots on the official WhatsApp Business Cloud API. Messages your business starts use templates approved by Meta, so reminder wording is written and submitted before launch (WhatsApp Cloud API).
DLT registration
Outbound calls and messages to Indian numbers run under DLT registration. We handle that as part of the build: registering your entity, your sender IDs and the message templates you want to use, then connecting them to the agent. We have done this on a project with outbound running under it. Approval of a template sits with the operators, so we do not put a date on it or promise a given template will pass.
The stack behind the calls
We run real-time voice on a self-hosted LiveKit setup for our own voice work. We use Plivo for Indian numbers (Plivo). Our sales team runs its own follow-up calls on Microsive Dialer, which we built (Microsive Dialer). If you want people rather than an AI agent making these calls, the Dialer gives reps a call list in the browser.
How we build it
You get a written scope and a fixed quote before work starts. Changes are quoted before they are done. You see the agent working at regular demos. The code and repository are yours at handover.
For calls coming in rather than going out, see inbound calls.
FAQ
What do you need from us to start?
The list of people to follow up with and why they expect to hear from you, the hours you want used, and access to wherever the outcomes should land.
Are follow-up calls recorded, and how are people told?
Yes, they are recorded. The agent says so at the start of the call, before it asks anything. The recordings sit in an account you own, beside the outcome of each attempt.
Which language will the agent speak?
The one your callers use. Language is a project decision rather than a fixed list: tell us who the agent will talk to, and we test a model in that language before we commit to it in the scope. That includes languages spoken outside India.
Which follow-up gets missed most?
A few lines on who you contact, and when, are enough to start. Someone from our team replies within one business day.