Time
1 month
Location
USA
Sector
Fintech
Description
The project’s objective was to create a scalable system for predicting financial market behavior by analyzing data from major global news agencies.
Solution
The solution involved the development of a data analysis system capable of processing data from various news sources, enabling the identification of correlations and causal links between significant global events. The project utilized approximately 20 servers and employed a distributed database along with a horizontally scalable micro-service architecture to facilitate flexible parallelization of calculations. Throughout the project, various challenges were encountered, such as addressing the issue of semantic unit similarity and optimizing load distribution across computational modules and the database to expedite data processing.
Results
The project resulted in the deployment of a system on the client’s side that enabled the identification of the five most profitable investments in financial instruments over a three-month period. This contributed to a 4% increase in the company’s revenue.