Case Studies

How Capital Group Transformed its Quant Trading Strategy with Automated News Aggregation and Sentiment Analysis

Enabled sentiment analysis capabilities that now allow enterprise leadership to track and respond to macro news events with confidence and agility.

Automated inefficient and resource-intensive manual research processes.

Challenge

Capital Group is an American financial services company with offices around the globe and over $2.6 trillion in assets under management. At the start of their engagement with Hakkoda, their team struggled with the overwhelming task of manually aggregating massive volumes of news stories and compiling a curated list of relevant articles for their CEO.

This process was essential to informing mission-critical operations, but required a full team and countless hours to complete. The firm also faced limitations when it came to using sentiment analysis in its development of quant trading strategies, combined with bottlenecks in their manual research processes that drained resources and delayed actionable insights.

Solution

Hakkoda automated Capital Group’s web-scraping process, efficiently aggregating news and media information from diverse sources and assigning scores to articles in order to automate recommendation for the company’s executive leadership. This eliminated the need for manual aggregation, saving the enterprise considerable time and resources. Leadership now receives a steady flow of comprehensive data, empowering timely and informed decision-making.

By implementing an agentic Streamlit application that leveraged AWS Bedrock’s natural language processing functionalities, Hakkoda enabled Capital Group to unlock sentiment analysis capabilities for their quant trading strategies. This advanced tool provided critical insights into market sentiment, allowing leadership to act swiftly on data-driven opportunities.

Hakkoda’s LLM-powered, automated solution now produces regular executive summaries and analyzes macroeconomic news sentiment. By replacing labor-intensive research processes with cost-efficient automation and leveraging natural language processing capabilities to make aggregated content easier to parse at scale, this solution significantly reduced the time required for executive leadership to gather insights while freeing up human resources to focus their efforts on other strategic priorities.

The Model

Technology Used:

Snowflake

AWS Bedrock

Streamlit

Microsoft Azure

Full Time Resources:

2 AI Engineers

Project Duration:

1 Month

Case studies

Case Studies

How an International Wellness Giant Modernized its Clinical Trial Screening Processes

Discover how an international wellness leader optimized its clinical trial screening processes using NLP and LLM capabilities in Snowflake.

Case Studies

How Hakkōda Enabled Brand Consolidation for a Leading Wealth Management Group

Learn how Hakkoda helped a leading wealth management group consolidate 11,000 advisors under a single brand umbrella while reducing monthly...

Case Studies

Unlocking Epic EHR Data for a Large Healthcare System

Learn how a large healthcare system integrated Epic EHR data with other data sources in a central cloud platform built...

Case Studies

Case Studies

How Capital Group Transformed its Quant Trading Strategy with Automated News Aggregation and Sentiment Analysis

Developed and deployed a news aggregation and curation solution using Streamlit and AWS Bedrock.

Enabled sentiment analysis capabilities that now allow enterprise leadership to track and respond to macro news events with confidence and agility.

Automated inefficient and resource-intensive manual research processes.

Challenge

Solution

The Model

Technology Used:

Snowflake

AWS Bedrock

Streamlit

Microsoft Azure

Full Time Resources:

2 AI Engineers

Project Duration:

1 Month

Case studies