Price: 0
Number of applications: 0
29.12.25 (inclusive)
it is considered individually
MVP
R&D Tasks
Robotics
Intelligent control systems
Software/ IS
With the growing volume of information uploaded by educational, scientific and government institutions, traditional monolithic databases are losing their ability to provide stable speed of search and analysis. Problems that need to be solved: • Lack of a unified text storage architecture at the national level; • a drop in query processing speed with an increase in the data corpus; • low efficiency of traditional indexing algorithms in the context of big data; • the need to ensure isolation of each organization's data while maintaining the possibility of centralized analysis; • The need for a stable and scalable computing engine for semantic proximity analysis and loan detection algorithms.
• creation of a technological architecture that allows processing search queries in milliseconds with data volumes of tens of millions of documents; • improving the accuracy of text loan analysis through deeper indexing and semantic processing algorithms; • reducing the load on the central infrastructure due to the distributed storage model; • the ability to scale the system to cover all educational institutions of the Republic of Kazakhstan and then expand to the CIS countries; • creating a technological foundation for the implementation of proprietary AI analytics algorithms and local models without dependence on external services.
Asylbekov Ulan
Purpose and description of task (project)
The goal of the project is to create a technological architecture for a distributed indexing database that provides long-term storage, fast processing, and scalable access to large amounts of text data (scientific articles, dissertations, books, publications, and online resources). The task provides for: • development of a data storage model based on distribution across organizations and regions; • implementation of algorithms for morphological normalization and semantic text processing; • Optimization of indexing to ensure high search speed even as the volume of data grows to the level of national archives; • Research on text proximity analysis algorithms, including shingles, vector representations, and neural models; • Designing mechanisms for horizontal scaling and fault tolerance.
Whatsapp
