CGI TextAI reduces workload and improves data analysis 

To gain a comprehensive understanding of employee sentiment, one of CGI’s business units used this solution to provide deep analysis of its annual internal consultation into CGI’s strategic direction. 

Challenge

Due to the large amount of data gathered for this survey, the company’s business unit faced three distinct challenges when analyzing the feedback:

  • Data overload: Manually sifting through and categorizing vast amounts of feedback was labor-intensive.
  • Topic diversity: The responses spanned topics not always explicitly asked in the survey questions, complicating the analysis. 
  • Need for comprehensive insights: Given the varied nature of the feedback, it was challenging to capture all the insights without overlooking critical details.

Solution

CGI TextAI’s capabilities helped to perform the survey analysis more efficiently and to realize deeper insights by using advanced machine learning (ML) and AI. The data analysis used several CGI TextAI features: 

  • Word cloud generation provided a visual representation of employee sentiment and responses—making it easier to quickly spot trends. 
  • Topic extraction grouped similar comments, helping to identify underlying themes.
  • Text summarization provided an overview of primary employee suggestions. 
  • Sentiment analysis uncovered insights into the overall positive, negative, or neutral feedback sentiment.
  • Clustering approach used ML to group similar comments and extracted topics without the requirement of external labels.
  • Topic distribution analyzed the distribution of topics across different questions. 
  • Flexible topic clustering merged related topics, guided the system to search for specific themes, and created interactive visualizations for more profound insights.
  • Pre-trained models were quick to use and adapted to the company’s specific data.

Outcomes

CGI TextAI drastically reduced the business unit’s manual workload, saving up to 30 person days by automating data analysis and extraction. Highly adaptable, the solution allowed for modifications to the analysis based on the business unit’s evolving needs. Pre-trained models also ensured the system could be quickly implemented and fine-tuned to ensure timely insights. The quality of data analysis also improved. 

CGI TextAI analyzes structured and unstructured data at-scale while improving data analytics and user experiences as well as reducing operational costs. This solution’s AI capabilities can be customized for unique datasets, infrastructure requirements and content needs across industries.