CallMarker Call Analysis with AI: Automatic Analysis of Every Call

In short: CallMarker already records and transcribes your call center's calls. One automation in CallMarker sends every transcribed call to TalkWiz when it ends, together with the agent's name, the customer, the campaign and the status. Within a minute you get a full analysis of the call: an agent score, outcome, objections, next step and alerts for the manager. Setup takes a few minutes and needs no developer.

What you get from the connection

• Analysis of every call, not a sample: a score by criteria, summary, objections, pain points, buying signals and competitors.
• Comparison with the CallMarker status: the status the agent marked (interested, not interested, follow-up, completed) is compared with what actually happened in the call.
• Breakdown by agent and campaign: because the agent's name and the campaign arrive with the call.
• A ready follow-up message for the customer on WhatsApp or email.
• A smart daily report and alerts about calls that need handling.
• The option to send the analysis back to CallMarker as a note on the customer.

Setting up the connection, step by step

1. Create a dedicated address in TalkWiz

In TalkWiz settings, under "Integrations", click to create an address for the automation. The address contains your organization's secret key, and it is shown only once. It can only send calls, not read data. If it leaks, create a new address and the old one stops working.

2. Create an automation in CallMarker

In CallMarker, create a new automation with these settings:

Setting	Value
Trigger	Call AI Analysis Finished
Method	POST to the address from the previous step
Request body	A field named transcript with the value [call:transcription]

3. Add the call details

So the analysis is linked to the agent and campaign, add more parameters to the same automation, in the request body or in the address. TalkWiz recognizes English and Hebrew names, for example:

• Agent Name: the agent's name
• Customer Name and Number: the customer's name and phone
• Campaign: the campaign
• Status: the status the agent marked
• Call duration, call date, call time: call duration and time
• Call ID: the call ID

A parameter that is not recognized is not lost: it is saved and shown on the call page.

4. Check that calls arrive

After the next call in the call center, the integrations screen in TalkWiz shows when the last call was received, how many calls were received and processed in the last 7 days, and which parameters arrived in the last request and which field each was matched to. If no calls arrive for 24 hours, an alert appears.

Mapping CallMarker statuses

In CallMarker, customer statuses include interested, not interested, follow-up and completed. TalkWiz uses the same terms for the call outcome, so they are easy to compare. If statuses arrive as numeric codes, give each code a name once in settings, and every report shows the name.

Good to know

• Calls whose transcription failed in CallMarker are not sent, because the automation runs only when the call's AI analysis finishes.
• Calls with no real conversation, for example a hang-up right after answering, are detected, not analyzed and not counted toward usage.
• The transcript is stored encrypted, and details such as credit card numbers are hidden before the analysis.

Next step

Open a free account, create the address and set up the automation. The first 100 calls are analyzed on us, with no credit card. Want to understand first what the analysis checks? Read the complete guide to AI call analysis.