How Bay Point Advisors Built an Intelligent Investor Portal with DataStax and Microsoft Azure
Bay Point Advisors is a private investment firm based in Atlanta, Georgia, focused on direct lending in underserved and non-traditional markets. With a legacy of building sophisticated financial tools and systems, Bay Point is now at the forefront of AI-driven innovation in the capital markets, leveraging new technologies to drive smarter investment decisions and enhance client engagement.
POC to working app in under two months
Reduced investor onboarding friction
Boost in analyst productivity
POC to working app in under two months
Reduced investor onboarding friction
Boost in analyst productivity
Challenge
Bay Point faced a growing challenge on how to scale operations and empower investors without sacrificing the personalized service high-net-worth clients expect. Traditional data rooms (repositories of hundreds of documents related to investments) were overwhelming and inefficient. In addition, internal teams struggle to derive fast insights from troves of unstructured loan documents, limiting their ability to identify new opportunities or respond quickly to customer queries. Bay Point knew generative AI had potential, but lacked a secure, enterprise-ready way to integrate it with sensitive financial data.
“When our CEO asked me to ‘use AI to search out additional business,’ my first reaction was skepticism. But I realized that came from not fully understanding the tools available. Once I discovered RAG, I saw how we could actually apply AI to our real-world business workflows.” —Rick Atkinson, CTO, Bay Point Advisors
Solution: Powered by DataStax and Microsoft
Bay Point partnered with DataStax and Microsoft in their mission to develop Atlas, an intelligent investor portal infused with Retrieval-Augmented Generation (RAG) capabilities. With Langflow for rapid prototyping, Astra DB for scalable vector storage, and Azure OpenAI for secure LLM interactions, Bay Point built a solution that allows investors to interact directly with their documents through a natural chat interface.
Key technologies:
- Astra DB: Served as the vector store to index and retrieve insights from thousands of unstructured documents.
- Langflow: Used to rapidly prototype and visualize the RAG architecture, enabling non-developers to build early POCs.
- Microsoft Azure OpenAI: Provided secure, enterprise-grade LLMs with contractual guarantees that customer data would not be used for training.
Bay Point designed a seamless pipeline:
- Documents are uploaded to Azure Storage.
- Text is extracted, chunked, and embedded into Astra DB using the .NET SDK and Data API.
- A master prompt orchestrates queries through Azure OpenAI, referencing only retrieved document chunks.
- Citations and links to source documents are included for transparency.
“DataStax made it easy for us to go from a proof of concept to a production-ready AI experience. Astra DB became the brain of our platform, helping us bring structure to unstructured data and build trust with both our investors and internal teams.” —Rick Atkinson, CTO, Bay Point Advisors
Deployment Challenges
While prototyping in Langflow was fast and intuitive, migrating to Bay Point’s .NET-based production environment introduced complexity. Translating Python-based workflows into C# required additional tooling and support. Early hurdles included:
- Misconfiguration of embedding vs. foundation models
- Manual JSON handling before the release of the .NET SDK
- Trial and error in prompt engineering and chunking strategies
- Limited internal familiarity with vector databases
Close collaboration with DataStax engineers helped overcome these challenges, with Astra DB’s Data API and SDK eventually streamlining the development process and reducing code complexity.
Impact
- Public facing data rooms enhanced with AI search
- Bay Point built from POC to working app in under two months, aided by Langflow and DataStax SDKs
- Reduced investor onboarding friction by replacing document overload with natural language Q&A
- Internal productivity boost for analysts reviewing multi-hundred-page loan agreements
- Improved trust and compliance via private LLM deployment on Azure and built-in citation for every AI response
Bay Point Advisors turned a leadership mandate “use AI to find more business” into a strategic advantage by building an intelligent investor experience rooted in secure, explainable, enterprise AI. With the combined power of DataStax and Microsoft, Bay Point has not only streamlined investor decision-making but also laid the foundation for future AI agents that can autonomously source deals, evaluate investments, and optimize fund performance.
