Tenda transforms sentiment analysis with generative AI and achieves 100% operational efficiency

Tenda's limitation, using a manual sentiment analysis process, prevents the company from obtaining quick and accurate insights into customer satisfaction, making it difficult to identify patterns, trends and critical areas for improvement.

In addition, the absence of a specialized AI tool compromises the ability to prioritize corrective actions and retention strategies, negatively impacting customer experience and business performance.

We created a solution that uses advanced AWS Generative AI models, leveraging their ability to understand complex textual patterns to analyze contributor comments and provide strategic insights. The solution architecture integrates robust AWS services to ensure scalability, security and efficiency.

Analysis occurs on two main fronts:

Multicategory classification:

• We use Amazon SageMaker to train and deploy custom models that automatically analyze and classify comments into multiple categories, based on specific criteria defined by the company.

• Amazon Bedrock with Raiku model is used to process and understand the context of comments, allowing categorization more accurate and based on advanced natural language.

Analytical report generation:

• Analysis results are stored in Amazon S3 and processed with Amazon Redshift, allowing the creation of detailed analytical dashboards.

• Based on these results, our solution automatically generates a report containing: Summary of positive and negative points identified in the comments.

• Comparison of results with previous research, highlighting trends.

• Recommendations for improvements, guiding strategic actions based on the insights obtained. This solution transforms large volumes of feedback into actionable strategic information, facilitating agile decision-making and improving the organizational climate, with the reliability and efficiency of AWS infrastructure.

There was an improvement in the efficiency of the analysis process, with automation allowing a faster and more accurate assessment of the feelings expressed by customers / employees.
The solution made the research result more realistic and improved, as:

• More areas / departments were able to take part in the research,  Before the solution, the HR team did not have enough people to carry out the task across the entire company.

• It was possible to increase the classification category

• The efficiency estimated by the client was 100%, as they would need to double the current team to achieve the same standard with AI.