How we automated lead qualification with Retell AI, Make.com and ClickUp

In our view, AI should support people's work, not replace it. We use automation to make work simpler and faster, so that people can spend their time on the tasks that need them.

In an earlier article we described how we automated email drafting for sales teams with Make.com, ClickUp and OpenAI. This time we are adding a new function: lead qualification by a voice AI agent. The agent follows a set conversation flow, analyses the call and logs the results directly in ClickUp. A sales agent can then give priority to the most engaged leads.

We use Retell AI for both outbound and inbound calls.

What is Retell AI?

Retell AI is a platform for building AI voice agents that handle phone calls and tasks such as booking appointments. Its latency is 600–800 ms, which keeps the conversation close to real time and natural.

Its main features:

• call transfer between agents,
• sentiment analysis, which gauges the caller's emotions,
• integration with CRM systems.

Retell AI is SOC 2 and HIPAA certified. The platform is used in healthcare, finance and customer service.

Why we chose Retell AI

Before choosing Retell AI, we reviewed what was on the market. We looked at several alternatives, including Bland AI, Voiceflow and Synthflow. The two finalists were Vapi and Retell AI: both have a similar interface, pay-as-you-go pricing and API integrations.

After testing, we chose Retell AI, mainly for two reasons:

1. SIP trunking support: a business can connect its own telephony provider, which lowers call costs. This means lower per-minute rates, better call quality and infrastructure that can handle a high volume of calls.
2. Free, built-in agent testing: Retell AI lets you test an agent by chat and by voice directly in its dashboard, which makes development and debugging easier.

Retell AI also integrates with Cal.com for booking meetings. With a few configuration changes, we automated meeting booking as part of lead handling.

Our call workflow in Retell AI

We handle both outbound and inbound calls, each with its own set flow.

1. Outbound calls

Lead qualification starts when a sales agent selects “Schedule AI Lead Phone Call” from the action menu in ClickUp.

(opens the full-size image)

Step by step:

1. ClickUp starts an automation that sends a webhook request to Make.com.
2. Make.com verifies the data:
   • checks that a phone number exists,
   • validates the phone number format,
   • retrieves the lead's details,
   • pulls the existing ClickUp comments to understand earlier interactions,
   • analyses the data with OpenAI to personalise the call script.
3. The call is registered in Retell AI, and the ClickUp status changes accordingly (e.g. “CALL REGISTERED” or “ERROR”).
4. Retell AI makes the call and sends real-time event updates back to Make.com.

(opens the full-size image)

After the call:

• Retell AI sends the transcript and the call outcome.
• OpenAI analyses the transcript and produces a structured JSON report.
• Additional data from Retell AI is added to the report.
• The final summary is saved as a ClickUp comment, so the sales agent sees clear next steps.

(opens the full-size image)

2. Inbound calls

Inbound calls are handled differently. Because we focus on outbound contact with leads, we only accept calls from leads we have already contacted.

Step by step:

1. When a lead calls back, Retell AI sends a webhook request to Make.com.
2. Make.com checks the number against the outbound call log.
3. The system retrieves the data about the previous conversation.
4. Retell AI receives the updated conversation context, so the AI agent can answer with it in mind.
5. After the call, the data goes through the same flow as after an outbound call, so it stays consistent.

(opens the full-size image)

Summary

Retell AI can be configured in detail through its dashboard, its API and SIP trunking. By connecting Retell AI, Make.com and ClickUp, we automated lead qualification: sales teams can focus on the most promising leads instead of spending time on unqualified ones.

Compliance for AI calls in Europe

AI voice calls, especially within the European Economic Area (EEA), must comply with the GDPR (General Data Protection Regulation) and local telecommunications law. Good practices:

• Consent to call recording: if calls are recorded, the lead must be clearly informed at the start of the call, and in some cases their consent is needed. A typical notice: “This call is being recorded for quality and compliance purposes. If you do not wish to be recorded, please let us know.”
• Legitimate interest or consent: under the GDPR, processing personal data (including recorded calls) needs a lawful basis. If recording is not strictly necessary, explicit consent is the safest approach.
• Right to be forgotten: leads have the right to ask for their data to be deleted, including call recordings. The system needs a way to handle such requests.
• Telling people they are talking to AI: under the GDPR, businesses must inform people when they are interacting with an AI system. Good practice is to offer a conversation with a person on request.
• “Do not call” registers: some EU countries keep opt-out registers for unsolicited calls. Check leads against these registers before calling them.
• Storage and security: stored transcripts and recordings must be protected with encryption and access controls, so that only authorised people can open them.
• Retention: companies should set a clear retention period for call recordings and transcripts, in line with the GDPR principle of data minimisation.

If you would like to see how a voice AI agent could qualify your leads, write to us at kontakt@proop.pl.