Frederik von der Heyden

AI Integration

PhoneIQ

Sales and support are on the phone all day, and none of it ends up in the system. PhoneIQ records calls with consent, transcribes them on your own server and summarizes what was agreed. Personal data is filtered out before any model sees the text. Currently in beta.

PhoneIQ

Case Study

!

Problem

Important agreements from phone calls live in heads and sticky notes. Off-the-shelf cloud solutions send customer voices unfiltered to providers outside Europe.

Solution

A pipeline that separates telephony, transcription and analysis: EU provider for the call, own server for speech to text, PII filter before every model. Consent and AI labeling are mandatory fields, not options.

Result

Call summaries and tasks in the system without customer data leaving Europe. The beta is running; the pattern transfers to video calls and voice messages.

What the solution does

01

Telephony From Europe

Calls run through a provider in the EU. Recordings are deleted there after 72 hours. No US provider for voice or personal data.

02

Consent That Stays Provable

Every consent is logged and cannot be altered afterwards. Anyone who does not agree is not recorded.

03

Filter, Then Analyze

Transcription locally, then the PII filter, only then the analysis. Summary, tasks and sentiment are derived from anonymized text.

04

AI Made Visible

Every result carries the model used and a note that it was generated by a machine. No silent blending into customer data.

Key figures

EU

Telephony and storage

72 h

Until recording deletion

PII

Filter before every analysis

Beta

Status

Tech stack

PythonNext.jsLocal speech recognitionEU telephonyOwn Server

Analyze Conversations Without Losing Data Protection?

PhoneIQ shows how telephony and AI fit together. Let's talk about your case.

Get in touch

Questions, ideas, or a specific project?

Get in touch