A new hotline experience awaits you. When you call the Swisscom hotline, you can now state your query orally. You no longer need a keypad in order to be transferred to the relevant officer. Using a voice and intention recognition system, also known as speech recognition, we can recognise your query and switch you through to the right officer more quickly.
For the purpose of training and further developing the speech recognition system, Swisscom needs voice recordings of how you formulate your voice commands (e.g. ‘I want to view my invoice. How can I do that?’).
The voice recordings will not be used for identification purposes, but will rather be exclusively processed for the training and further development of our voice hotline solution.
Swisscom also records conversations between you and our officers for training and quality assurance purposes. These recordings allow us to train our officers and develop their conversation skills further.
Currently, our German-language customers have the benefit of the Swisscom hotline speech recognition system. The existing solution is being constantly expanded. Swisscom updates customers on an ongoing basis when new languages are added.
We record your voice commands and query descriptions (e.g. ‘I have a question concerning my mobile phone contract.’) These voice recordings are processed to train and further develop a speech recognition system. The processing includes the saving, analysis and modification of the voice recordings. The voice recordings are processed by Swisscom exclusively in Switzerland as well as by selected partners in Switzerland and in the EU.
The processing of voice recordings for development purposes also entails listening by humans. After training, the voice hotline solution should, based on the oral statement, be able to transfer a caller to the relevant hotline officer.
The voice recordings of the conversation between the customer and the officer are also used for training and quality assurance purposes.
A speech recognition system is software that translates language into text. Such systems are applied amongst others to automatic speech recognition, the translation of spoken language into written text or the translation between two languages. The aim of a speech recognition system is to interpret the meaning of the spoken content in order to carry out the requested actions.
When the Swisscom hotline receives a call, spoken voice commands and query descriptions are transmitted to the hotline platform and recorded. These voice recordings are processed in a speech recognition system to enable your voice commands to be executed.
The voice recordings are also processed in order to improve and further develop the Swisscom hotline speech recognition system. The processing includes, amongst other things, the collection, saving, management, use, transfer, deletion of, and access to, voice recordings.
When the recordings are not subject to a deletion request, the voice recordings are saved for five years at most. If the customer relationship with Swisscom comes to an end earlier, the voice recordings shall be deleted.
Your query statement will in that case be used only to transfer you to the appropriate officer, and will not be recorded in order to improve the voice system.
No, Swisscom will not create a user profile with your voice recordings. The voice recordings will not be used either for identification purposes or for marketing purposes. The voice commands are merely recorded and further processed to enable your voice commands to be executed and in order to improve and further develop the voice hotline solution. The recordings of your conversation with the officer are also used for training and quality assurance purposes.
Your voice recordings are used to improve and further develop the Swisscom hotline speech recognition system, as well as for the purpose of training and educating our officers.
When the Swisscom hotline receives a call, Swisscom records your voice commands and query statements or your conversation with our officers. This is used to train the speech recognition system or to develop our officers’ conversation skills. Should your call be recorded, you will be informed accordingly and be given the option of preventing a recording through simply pressing a key. In addition, should you have previously notified Swisscom of your opposition to voice recordings, this will be complied with.
In addition to automated transcription (translation of the recording into text), individual voice recordings may also be listened to, transcribed and classified. or compared to automated transcriptions, by Swisscom officers. This serves the purpose of improving and further developing the speech recognition system. The voice recordings are also used to train our officers and improve their conversation skills. To this end, voice recordings may be listened to by the officers who participated in the conversation and by those training them.
Voice recordings can be processed by contracted third parties (‘service providers’). Service providers provide automated transcription services and speech recognition solutions. Service providers who process voice recordings have their registered offices in Switzerland and the EU. They are under a duty to process the voice recordings exclusively in accordance with Swisscom instructions and to permanently delete any received voice recordings as soon as the contracted processing has been completed.
No, the transcripts are not anonymised. The data is, however, saved without any direct references to you (e.g. account number, customer name, etc.). In order to be able to respond to customer requests to access their information or to have it deleted, the voice recordings are allocated identification keys. These consist of a long sequence of digits which on its own is not sufficient to identify the customer. Only a small group of people is, at the customer’s request and to protect his/her rights as a data subject, capable of generating this identification key. This allows the voice recordings belonging to the customer to be identified and, if necessary, deleted.
The voice recordings are processed by Swisscom in Switzerland as well as by selected cloud providers on servers in Switzerland and the EU.
If the recordings are not subject to a deletion request, the voice recordings are stored for five years at most. If the customer relationship with Swisscom comes to an end earlier, the voice recordings shall be deleted.
When you call the Swisscom hotline, you will hear an announcement regarding the voice recordings. You will be informed that your voice commands and query statements, as well as your conversation with our officers, will be recorded and processed for further development and quality assurance purposes. If you object to this, you will be given the option of preventing the storage and use of a recording of your voice by pressing 9. You will hear the following announcement:
‘This call will be recorded for voice recognition improvement and quality assurance purposes. You are helping us to provide improved services to our customers. If you do not wish us to use the recording, please press “9”. Further information is available from our officers or at swisscom.ch/recording.’
You also have the option of adapting your consent to voice recordings through your customer center: www.swisscom.ch/voice-recording
Yes, you may have your voice recording deleted at any time later. You can submit your request to us in writing, enclosing a copy of your ID card or your passport, to our postal address:
Swisscom (Schweiz) AG
Legal & Regulatory
Datenschutz
Postfach
3050 Bern
Auskunft.Datenschutz@swisscom.com
In this case, your recording will no longer be used for training the speech recognition system and will be deleted.
Further information is available in Swisscom's General Privacy Policy (Section 8).
No, both the voice recording itself and the transcripts will be deleted.
Swisscom takes the responsible and legally compliant use of your personal data very seriously. Swisscom takes appropriate technical and organisational measures against unauthorised processing of the voice recordings. In particular, voice recordings are always encrypted and saved without any direct reference to you, such as your name or account number. The saved voice recordings will only be processed in relation to the relevant individual in order to comply with your rights as a data subject and only by a small number of persons entrusted with this task. This may be done in order to provide you with information regarding the voice recordings of you that exist or to delete these at your request.
All persons capable of accessing the voice recordings are under a duty of confidentiality and are obligated to comply with appropriate security measures.