Building a Scalable Document Querying Chatbot: Corbin Capital

Building a Scalable Document Querying Chatbot: Corbin Capital

Business Problem

  • Corbin Capital has implemented a Document Querying Chatbot to facilitate efficient and accurate document querying within its repository. This advanced solution leverages state-of-the-art NLP models and is designed to provide quick access to relevant information, significantly enhancing productivity and decision-making processes. Managing and retrieving data from vast document repositories can be time-consuming and prone to errors in large organizations like Corbin Capital.
  • Employees often need help finding specific information quickly, leading to delays in decision-making and reduced operational efficiency. The existing manual search processes could have been faster and more efficient, making retrieving precise information from many documents more accessible. The system could also have been more user-friendly, resulting in low adoption rates. Ensuring data security and regulatory compliance during information retrieval was another critical challenge, along with the high resource allocation required for training and manual searches.

About Corbin Capital

Corbin Capital Partners is a woman-led investment firm specializing in alternative assets like hedge funds and credit investments. They manage client money through various methods and prioritize client satisfaction. As of April 1, 2024, they manage $9.1 billion in assets.

Solution

Corbin Capital addressed these challenges by implementing a Document Querying Chatbot, an AI-driven solution designed to streamline information retrieval. The chatbot provides several key features to enhance the user experience and improve operational efficiency.

Building a Scalable Document Querying Chatbot: Corbin Capital
Click to View in Full Size

Architecture

  • User Interaction: Users interact with the system through a React.js user interface, submitting queries and accessing information securely via SSL certificates installed on a Linux VM.
  • Data Ingestion and Processing: Data from sources like GitHub and other repositories are ingested, processed by a Vision Parser, and stored as raw and processed files.
  • AI-Powered Services: The application utilizes Azure AI services (OpenAI Vision, Turbo, Embed, and AI Search) for advanced data processing, analysis, and efficient querying.
  • Data Storage and Management: Azure Cosmos DB stores and retrieves structured data, while an App for SharePoint integrates for enhanced document management and collaboration.
  • Security and Compliance: User authentication and authorization ensure secure access, with user interactions and feedback logged for continuous improvement and auditing.
  • Scalability and Efficiency: The architecture supports scalable and automated processes, managed by a scheduler to ensure timely task execution and efficient resource utilization.
  • Continuous Improvement: Logged user interactions and feedback are analyzed to refine and enhance the chatbot’s performance over time, ensuring it adapts to evolving user needs.
Outcomes

0Faster Answers, Better Decisions

Chatbot provides quick access to info, streamlining workflows and empowering informed choices.

0Shared Knowledge, Stronger Teams

Breaks down information silos, fostering knowledge sharing and collaboration across teams.

0Reduced Errors, Increased Accuracy

Eliminates human error in document searches, leading to more reliable data and accurate decisions.

Technology Stack​