FAQ

Frequently Asked Questions

We know you might have questions, therefore we collected Frequently Asked Questions about the conversational AI here.

Please feel free to contact us, if you want to know more:

What is Feelingstream?

Feelingstream is a conversation analytics tool that you can use to transcribe and classify conversations, and which provides exciting visibility. It can be used to discover new sales leads, detect conversations indicating customer churn but also bottlenecks in the processes or errors in customer service. Keeping existing customers happy is much more cost-efficient than trying to find new ones. Feelingstream helps companies stop losing money in customer service.

Who is Feelingstream for?

Feelingstream is the best conversation analytics tool for managers responsible for efficiency, customer service quality, or sales revenue. Various department heads get data for large scale business decision making, Quality Managers information for call analysis and agent training, Product Owners receive feedback for product and process improvements. The tool is useful for managers of all levels and various interest but can even be used by Agents to analyse their own conversations.

What problems does Feelingstream address?

Our users have faced the following challenges before starting to use Feelingstream : 
1. How to reduce call volumes and save Agents’ time  
2. How to ensure data quality in CRM in notes which Agents insert after every single call manually and spend time on    
3. How to sort out which calls to evaluate for service quality assessing purposes.  
4. Unscalable coaching for Agents who have just started their call centre career.   
5. Moving customers from calls to other channels such as self-service and website 
6. Email routing and automatic replies for efficiency 
 
There are multiple other use cases for our product, read more about the benefits in sales, quality, and efficiency. 

Which channels can be analysed with Feelingstream? 

Feelingstream product is an end-to-end platform for all conversational channels (audio, email, chat, NPS) which enables to evaluate customer conversations in real-time (in case of audio in near real-time) and react if targeted events are appearing. 
 

How accurate is the ASR (Automated Speech Recognition) model? 

Feelingstream adopted our ASR model specifically for customer service calls. This speech is well known as spontaneous speech (read about using speech to text for finding business-critical calls). We have achieved 85-90% of accuracy in various languages (for example, Finnish), which is a really high result.  
There are always some mistakes when doing automatic transcriptions and we work hard to improve our model based on these issues, but already today the accuracy is enough for analysis and automation of processes.   

How are the integrations done? 

All calls, emails or chats fly in to Feelingstream platform automatically through integrations. No manual uploading is needed. After the data is visible in the Feelingstream product, you can log in and start to consume the value. Yes, it works already from Day 1! We offer setup as a cloud or an on-premise service. Read more about our newest integrations

How does Feelingstream ensure the security of data? 

With Feelingstream, access during preparation and work is granted based on real needs and given only for time period needed model building/analysis. If personal data is not necessary, we can use anonymized data.  
We have implemented a granular user-groups based access policy. Users are entered into groups that have access based on their needs. We can anonymize conversations and change the human voice.  
All data transmissions are encrypted using industry-standard SSL (Secure Sockets Layer). Our mission critical systems have technical security layers including firewalls, authentication points and segregated networks. 


Did we answer your question? If not, please let us know by sending us an email: info[@]feelingstream.ai.