Recognizing polling questions from a conference call discussion

A non-transitory computer readable storage medium comprising instructions for a server that, when executed by a processing device, cause the processing device to perform operations comprising:

• obtaining one or more audio files comprising verbal phrases provided by a plurality of participants of a conference call;
• converting content of the one or more audio files into a set of text strings comprising a textual form of the verbal phrases;
• providing the set of text strings as input to a machine learning model, wherein the machine learning model is trained to predict, based on given textual data, a verbal phrase indicated by text strings of the textual data that comprises one or more questions for polling conference call participants;
• obtaining one or more outputs from the machine learning model;
• extracting, from the one or more outputs, a level of confidence that a verbal phrase provided by a first participant of the plurality of participants comprises a question associated with polling other participants during the conference call;
• responsive to determining that the level of confidence satisfies a confidence criterion, designating the verbal phrase as a polling question presented during the conference call;
• and
• causing the verbal phrase to be displayed in a user interface (UI) of a first client device associated with the first participant of the conference call, wherein the UI of the first client device comprises one or more UI elements configured to enable the first participant to pose the verbal phrase as the polling question to one or more additional participants of the conference call, and to edit at least a portion of the polling question.

A system comprising:

• a memory device;
• and
• a processing device coupled to the memory device, the processing device to perform operations comprising:
• obtaining one or more audio files comprising verbal phrases provided by a plurality of participants of a conference call;
• converting content of the one or more audio files into a set of text strings comprising a textual form of the verbal phrases;
• providing the set of text strings as input to a machine learning model, wherein the machine learning model is trained to predict, based on given textual data, a verbal phrase indicated by text strings of the textual data that comprises one or more questions for polling conference call participants;
• obtaining one or more outputs from the machine learning model;
• extracting, from the one or more outputs, a level of confidence that a verbal phrase provided by a first participant of the plurality of participants comprises a question associated with polling other participants during the conference call;
• responsive to determining that the level of confidence satisfies a confidence criterion, designating the verbal phrase as a polling question presented during the conference call;
• and
• causing the verbal phrase to be displayed in a user interface (UI) of a first client device associated with the first participant of the conference call, wherein the UI of the first client device comprises one or more UI elements configured to enable the first participant to pose the verbal phrase as the polling question to one or more additional participants of the conference call, and to edit at least a portion of the polling question.