Devoteam is responsible for providing the advanced AI technology necessary for the project. By leveraging Google’s Gemini, they offer a robust framework that enhances both the summarisation of news articles and the subsequent sentiment analysis.
About the customer
Katadata is an online media and research company in the fields of economics and business that combines the power of journalism with sharp analysis. The company provides a variety of credible information and data about the Indonesian economy, business, and finance as a reference source for business practitioners, investors, and policymakers. Some of the research conducted by Katadata has focused on the upstream oil and gas industry, mining, and banking. These studies are published and subsequently serve as reference material for mass media reporting.
The Challenges
- Manual review inefficiencies: Current process of reviewing comments and feedback is time-consuming and misses valuable insights.
- Scalability issues: Existing system using OpenAI’s ChatGPT struggles to handle the growing volume of news articles and sentiment data.
- Accuracy concerns: Inconsistent sentiment analysis affects the relevance of information provided to the audience.
· Market opportunity:
- Lack of sentiment analysis case studies in Indonesia creates a gap, offering Katadata a chance to lead in this space.
· Solution needed:
- AI-driven framework: Katadata requires an innovative, robust framework to improve news summarization and sentiment analysis.
· Goal:
- Katadata aims to elevate the quality of its information, reinforcing its reputation as a trusted source for business professionals, investors, and policymakers.
The Solutions
In the rapidly evolving digital landscape, Devoteam has introduced an advanced solution leveraging Google’s Gemini. This innovative system is designed to significantly enhance the summarization and analysis of news articles, providing insightful evaluations of public narrative. Gemini’s capabilities are pivotal in this solution, as it not only enhances the speed of summarization but also improves the accuracy of semantic search evaluations. By leveraging advanced natural language processing and machine learning algorithms, Gemini ensures that the insights derived are both timely and relevant, providing audiences with a deeper understanding of public narratives based on credible data sources.
The solution employs a two-phase process to ensure comprehensive coverage and accurate insights:
- Phase 1: News Article Summary The process begins with skilled writers crafting news articles. Subsequently, AI technology generates concise summaries, distilling essential information and data for further analysis.
- Phase 2: Semantic Search As a solution to enhance the relevance of search results on the Katadata and Databoks platforms, Semantic Search can be implemented. This technology utilizes machine learning models to understand the context of the keywords entered by users, delivering more accurate and relevant search results. By incorporating semantic search, users will be able to more easily find articles, reports, and critical data based on the meaning of the words, not just literal keyword matches.
The importance of AI in Semantic Search AI-driven semantic search enhances the accuracy and relevance of search results by understanding the context of queries, not just keywords. This allows users to quickly find the most relevant content, improving the overall search experience and decision-making process.
The Results
Generative AI, powered by the Flash model, offers a streamlined approach to summarizing content, allowing users to quickly extract key points without unnecessary elaboration. This advanced model efficiently generates concise summaries, helping users save up to 30% of their time, thereby boosting productivity.
Moreover, summaries created by one user can be saved and accessed by others, enhancing overall system efficiency by up to 40%. This collaborative feature promotes knowledge sharing, as teams can leverage insights from their peers, leading to a 15% improvement in team performance.
The Flash model also focuses on summarizing only the articles that have been opened, optimizing resource usage and avoiding redundant processing. This targeted approach ensures that the AI prioritizes relevant content, making the summarization process both cost-effective and efficient. As a result, organizations can reduce operational costs related to information processing by 20%, streamlining workflows and maximizing business impact.
In conclusion, utilizing Generative AI with the Flash model offers significant gains in efficiency, collaboration, and cost savings, making it an invaluable solution for businesses aiming to optimize workflows and enhance productivity.
“As a user of Devoteam’s services, we at Katadata truly appreciate their affordable solutions, backed by a highly responsive and solution-oriented team, ensuring our needs are always met quickly and efficiently.”
Anugerah Trihatmojo,
CTO at KATADATA
Your Success Starts Here
Partner with Devoteam to leverage award-winning tech expertise, agile execution, and a culture of continuous learning.