Inspiration
As a summer student at an insurance company, I have been tasked with classifying the reasons for old claims and deriving insights to improve service delivery. Inspired by this challenge, I plan to build an Analysis Assistant to streamline and enhance the business analysis process. Although I have completed Phase 1, time constraints have prevented me from finishing Phase 2. Nevertheless, I intend to continue this project as a side endeavor, utilizing the knowledge gained from this learning hackathon. I appreciate the opportunity provided by this event and Microsoft's support.
What it does
The Analysis Assistant is designed to assist in the classification of old insurance claims and provide actionable insights. It takes an Excel file containing text that needs classification and uses AI to process and categorize the data. Additionally, it integrates with a document or PDF repository stored in vCore-based Azure Cosmos DB, enabling the assistant to reference insurance policy terms and conditions. By writing a prompt, users can instruct the AI on how to perform the classification and search for relevant information in the DB using RAG. The AI processes the text from the Excel file, searches the DB and the n generates the good prompt, and returns the classified data in a new Excel file.
How we built it
Frontend: Built a user-friendly web interface for users to upload Excel files and documents. Backend: Utilized Python for data processing and classification. Database: Used vCore-based Azure Cosmos DB to store and manage insurance policy documents. AI Integration: Open AI api
Challenges we ran into
Data Integration: Integrating various data sources, such as Excel files and Cosmos DB documents, was challenging. Time Constraints:
Accomplishments that we're proud of
Phase 1 Completion: Completed Phase 1, laying a solid foundation for the project. Phase 2: Build a good idea for a side project and help my work.
What we learned
Learned how to combine vCore-based Azure Cosmos DB for MongoDB vector search and document retrieval with Azure OpenAI services to build a chatbot.
What's next for AnanysisAssistant
Phase 2 Development: Continue working on Phase 2 as a side project.
User Testing: Conduct user testing to gather feedback and make necessary improvements.
Feature Expansion: Add new features, such as more advanced analytics and reporting capabilities, to further assist business analysis.
E-commerce Integration: Integrate with an e-commerce website to allow users to register, log in, and purchase points for using the Analysis Assistant.
Frontend: Develop a user-friendly interface using React.
Backend: Implement core functionalities with Python Django.
User Management: Use Java Spring Boot for handling user registration and login.
Cart and Store: Utilize Java Spring Boot to manage the shopping cart and store operations.
Microservices Architecture: Transition the system to a microservices architecture to improve scalability and maintainability, ensuring each component can be developed, deployed, and scaled independently.
Log in or sign up for Devpost to join the conversation.