Financial Analysis Assistant

FAA or Financial Analysis Assistant is a steppingstone to help the stock analyst and/or trader community to get real-time, accurate and relevant information about the companies that they are interested in. With the help of AI, the user community would be able to quickly and accurately search through mountains of public data and formulate the right questions that need to be answered to make a sound investment decision.

In fact, the current version of the assistant is so easy to use that anyone can benefit from it without the need to have extensive financial markets expertise. I believe that the tool can actually be used in schools from 5-8 grades to teach the students the fundamentals of investing and financial capital market concepts. This can be a great tool for educators.

Alternative approach can be in two versions for two different user communities:

1) Non-specialized users (beginners) in finance

2) Specialized users (experts) in finance, e.g. stock analysts, traders and portfolio managers.

1) For non-specialized user base, the FAA can be deployed as a Web page but with underlying development to further train the models into producing more accurate, precise information to the user with extensive prompt engineering. This prompt engineering would further facilitate and simplify user interaction with FAA. Also, users would be able to access similar information in this version with richer User Interface and with additional features like speech recognition. The financial data retrieval and search can also be improved to use unstructured company data from the Web. This version can be used in schools and train young minds to learn about capital markets. A more advanced version of this tool can also be deployed to XR headsets like Meta Quest or Apple Vision Pro for greater impact.

2) For specialized user base, the alternative approach is to develop a piece of software with which stock analysts can use to quickly decipher company CEO and senior management comments in real-time and formulate meaningful and relevant questions during quarterly earnings calls that would refer back to a specific past comment or financial data. The financial data that FAA would be searching through is stored as unstructured data in public domains like SEC or company sites. The tool would ultimately show the financial data in graphical format and be implemented on AWS Bedrock and deployed to XR headsets like Meta Quest or Apple Vision Pro.

Finally, in both alternative approaches, vector search and sophisticated RAGs will be used in varying degrees to tend to software audiences.

Please see the AWS Bedrock Partyrock Hackathon entry details here: https://devpost.com/software/financial-analysis-assistant