Challenges faced

The client specializes in providing Business Intelligence to other companies that are looking to expand their business internationally. They used to do this by going through a plethora of data sources like business filings, articles, CEO interviews, social media posts, news, etc. Their team manually compiled reports that provided insight into risks, challenges, and optimization opportunities for other companies striving to expand their businesses. This manual exercise basically entailed the following problems:

  • Labour intensive:

    One of the major challenges faced by the client was to deploy an enormous workforce, which ultimately led to high production costs.

  • Slow:

    Since there was no automatic workflow, the entire process of transformation of raw documents to a proper readable format was tedious and took a substantial period of time.

Proposed Solution

TechVariable developed a web-based tool for Market Research report generation from various Government documents and other sources like social media, news, articles, etc., involving extraction, analysis, and presentation in a readable as well as organized document. Our approach was to go with Microservices architecture that could be expanded over time. The project was divided into multiple releases. Release 1 entailed using 3rd party APIs to aggregate different sources of news as well as social media posts. This release included a keyword search feature as well, available from government filing forms and documents. Also, stock related information was also integrated within the application. An email notification system was also developed along with the integration of clickstream analytics.

Release 2 basically was to develop a tool for the collection of data from the quarter and annual filings released by SEC (Securities and Exchange Commission). Parsing through the filing data, our NLP algorithm selectively picked out sentences within the vicinity of contexts relating to international expansion, and the NLG portion compiled these data into a presentable and human-readable format. Release 3 was developed to showcase investor presentations within the developed interface. Release 4 basically dealt with a similar kind of tool as Release 2, but this time the data source was DEF14A forms and there was an added need to compartmentalize the aggregated data into different headings prescribed by the client. Release 5 dealt with the development of earning transcripts (alpha transcripts) interface.