InvestorMegamind.AI

Investor_Relations_Agent_Workflow
test script part 1
test script part 2
test script part 3
test script part 5
test script part 4

Inspiration

The primary inspiration behind creating the AI agent was to help novice investors gain clarity, get answers to their questions, and make informed investment decisions. The agent provides invaluable support and guidance, empowering investors with the knowledge they need to approach investments with confidence. For organizations aiming to educate investors, the AI agent offers tremendous value through its reinforced learning capabilities, providing personalized and relevant information. The Investor Relations AI agent bridges the knowledge gap for novice investors and provides organizations with a powerful tool to effectively educate and support their investor community.

What it does

Provides useful information about Berkshire Hathaway's business operations, real-time stock data, and upcoming events.
Offers support and guidance, empowering investors with knowledge to approach investments confidently.

How we built it

We leveraged Berkshire Hathaway's annual shareholder letters for domain knowledge and stored them in a Data Store. This helped with the augmentation of knowledge for the generative model.
Leveraged Berkshire Hathaway's annual shareholder letters for domain knowledge (stored in Data Store).
Utilized Polygon Finance API for real-time data and Google Cloud Functions for events data.
Watched YouTube videos and referred to Google Cloud documentation to understand Agent Builder.

Challenges we ran into

Integrating agent into external integrations and interfaces, including avatars.
Navigating documentation and finding real-world examples.
Insufficient documentation on certain topics and lack of proper error responses.

Accomplishments

Created a functioning agent that accomplishes our goals.
Gained understanding of end-to-end AI agent building using internal and external data, creating a production-like agent in 2 days using low-code tools.

What we learned

Leveraged tools like Data Stores and external APIs in AI agent builder.
Discovered a no-code way of building an agent by augmenting knowledge and grounding results.

What's next for InvestorMegamind.AI

Iterate on the idea and build the product with easy integrations.
Approach investor relations departments to pitch and integrate the product into their portals.