Sentiment Analysis

Analyze and measure a speaker’s expressed feeling or enthusiasm about a specific topic using Sentiment API


High accuracy

The AI model behind the Sentiment API is designed to handle the complexities of human language and can accurately identify the sentiment of human conversations with high accuracy

Convenient use

Compared to other models, Symbl’s model limits the API calls you have to make if you require more than just sentiments increasing ease of use.

Aspect based sentiment analysis

Sentiment analysis performed in conjunction with topic modeling (through Topic API) so as to capture contextual relevance of the sentiment within topic as well as broader conversation


Extract sentence level sentiment polarity to understand emotional context in mission critical conversations.

Multiple inputs

Sentiments are available for audio, video and text inputs.

Global polarity labels

Sentiment is provided as a polarity in the range -1 to +1 along with a suggested label from positive, neutral, and negative.

Use Cases

Customer Satisfaction

Understand speaker engagement, talk time, silence, and sentiment to analyze customer satisfaction and take action to improve outcomes of sales and customer care calls.

Agent Performance

Analyze the conversation dynamics of top performing customer care agents or sales reps and create a repeatable recipe for success.

Realtime Call Mitigation

Detect when calls are at-risk or customer satisfaction is falling and create agent-assist workflows to adapt the call-flow in real time.


Processing conversations

Start by processing real-time or pre-recorded conversations using the Streaming API and Async API.

Perform Topic level sentiment analysis

Identify members and topics in conversation and analyze messages and factors for each topic.

Set custom definitions

Fine-tune sentiment polarity labels and scores to cater to your unique business requirements.