Conversation Analytics
Understand your customers in real time! Our AI (Artificial Intelligence) is able to understand single words and complete sentences for a correct analysis of the chat conversation and a subsequent classification according to the subject: cancellation, post-sale, claim, etc.
Our tool automatically labels each comment identifying the topic mentioned with its respective sentiment.
If a comment has several topics, it is multi-tagged allowing a comprehensive understanding of the topics mentioned.
Discover and understand what consumers really want
Our tool makes it possible to understand your consumers intentions in real time, how well you’re fulfilling them, and those that can be easily automated
Start with the intent, end with results for your business
Keepcon uses our industry-leading NLU engine to detect consumer intents, identify conversations to automate, and inform critical business decisions.
Keepcon transforms how you run your business
Understand consumer intents and automate conversations
Detect consumer intents in real time with our proprietary NLU engine and automate them with conversation flows. We’ve used over a billion conversations to train machine-learning algorithms that power our AI. Our NLU outperforms many competitive benchmarks related to accuracy, precision, and recall across a variety of industries.
Optimize contact center operations with intent-based performance data
Identify intents with poor consumer sentiment and take action. Route intents to the bot or agent best equipped to resolve them. Improve agent training and tune bot performance to improve intents with low sentiment.
Easily map new, misunderstood phrases to existing intents
Turn human insights into actionable feedback to improve intent comprehension with Logios. This makes labeling conversation data and providing performance feedback stress-free — helping agents be more productive and bots a lot smarter.
Get up and running quickly with intent starter packs for your industry
Logios, machine-learning algorithms, and deep learning neural networks analyzed over a billion conversations to classify the top intents for a variety of industries. Proprietary large-scale, pre-trained language models (ELMo) were used to preconfigure these intents. This makes it possible for Keepcon to automatically recognize up to 65% of intents with little-to-no configuration
Successful cases
Sales process | Case 1
Objective: Increase cross-selling, verifying the moments of the process where most of the potential sales fall.
Results after making decisions based on deep analysis of conversations:
- 30% increase in cross-sales
- ROI: 6 months
- 50% reduction of the audit team
- Compliance risk reduction
Collections process | Case 2
Objective: Increase the acceptance rate of a payment agreement by the client, which has a direct impact on the collection rate.
Results after making decisions based on deep analysis of conversations:
- Increase of between 9% and 17% in the collection rate
- Multiplication x3 of debt recoveries
- These results were achieved by maintaining the satisfaction indexes