AI-Based News Analytics Solution for Hyper-Personalized Investment Recommendations

AI-Based News Analytics Solution for Hyper-Personalized Investment Recommendations

John David

The client is an investment advisory firm that serves corporations, governments, institutions, and private clients globally. It aims to help its client’s financial well-being through tailored money management, industry-leading service, global investment diversification, and illuminating insights. They intended to automate their manual research process of qualitative analysis of stocks, mutual funds, and ETFs to automate their investment advice, guidance, and personalized reporting features.

Client: Banking & Finance

Services: AI, ML, Data Science

Year: 2022

Key Challenges

  • The existing reporting and interface systems lacked credibility, and not all the reports pushed to investors were very useful in taking action on one’s investments​​​​​​​​​
  • Faced challenges in performing manual research for qualitative analysis of stocks, Mutual funds, and ETFs ​
  • The client needed a solution that could automate their investment advice process and offer better, personalized, and relevant information to investors to help meet their financial needs ​
  • The client wanted to address their mobile application inefficiencies and needed a very simple UI/UX, so that they wouldn’t become overwhelmed and confused while navigating through the app ​

Solution

  • Sparity implemented a standalone News Analytics solution that offers investors detailed reports, risk factors, personalized news articles, and relevant investment information so that investors can make informed financial decisions
  • Leveraged AI and ML to implement data scraping process from various internal & external sources and text analytics algorithms to identify macro level insights and trends ​
  • Sparity employed Java Spring Boot with Microservices architecture for backend ​ ​​​​
  • Used Bloc architecture in mobile apps using flutter and deployed to both Android and iOS app stores ​ ​
  • Performed Text scraping from financial websites (Nasdaq news, Stocktwits) to extract sentiment using Python and Selenium ​
  • Employed Selenium along with the BeautifulSoup library for scraping the NASDAQ news ​
  • Employed pandas library for constructing a simple data frame from the scraped information to organize the collected data in a table and view the stored results ​​
  • Visitor Segmentation- Profile and segment customers based on Weblog/ CRM/ Sales data ​
  • Visitor Path Analytics-Visitor behavior can be understood through web visit paths to recognize patterns​
  • Affinity Analysis- Creating correlation between products/Reports /news Items and customer persona and portfolio ​

Benefits

  • Intuitive engagement platform delivers ease of access with rich featured user interface
  • Delivered more accurate and relevant investment information results in high levels of client satisfaction ​
  • 10x improvement in meeting investor’s Investment objectives
  • Increased operational efficiency and found an operating cost reduction of 45%
  • 70% increase in both iOS and Android app users