Project Goals
Secure Document Vault
Implement a secure document vault system for storing and organizing company documents
AI-Powered Document Querying
Allow users to interact with AI models for querying documents and gaining insights in real-time
Role-Based Access Control
Provide role-based access control for managing user permissions within different operational domains (Quality, Safety, Compliance, Troubleshooting)
Customizable Chat Interfaces
Enable customizable chat interfaces tailored to each operational domain to assist with user queries
Secure Authentication with Keycloak
Integrate Keycloak authentication to ensure secure access for users


Challenges
Complex Document Management
Managing large volumes of operational documents in a way that was secure, organized, and easily accessible was a major challenge.
AI Integration
Integrating AI models (OpenAI GPT-4o-mini and Google Gemini) to provide contextual document insights and generate accurate responses required careful configuration and testing
User Access and Security
The platform required a robust role-based access control system to ensure that the right users had the right level of access to sensitive documents.
Real-Time Interactivity
Users needed to query the AI models in real-time and receive quick, actionable insights without delays.
Solutions
React.js
Developed the frontend with React.js, enabling a dynamic, responsive user interface. This allowed users to interact with the platform in real-time, manage documents, and engage with the AI chat system seamlessly
Nest.js
Powered the backend API using Nest.js, providing a robust architecture for handling complex data flows, user authentication, document management, and AI interactions. Nest.js ensured that the application could scale efficiently and integrate seamlessly with AI models
MongoDB
Used MongoDB as the database for secure document storage, allowing for flexible and scalable management of both structured and unstructured data. The database also supported efficient querying and retrieval of documents based on AI interactions
AI Models (OpenAI GPT-4o-mini & Google Gemini)
Integrated OpenAI GPT-4o-mini and Google Gemini to provide AI-powered content generation, document analysis, and intelligent responses to user queries


Conclusion
The Nestlé AI-powered knowledge management platform successfully transformed Nestlé’s internal operations by enabling faster, smarter decision-making. The platform, built with React.js, Nest.js, and MongoDB, provided secure document management, AI-driven insights, and seamless communication across operational domains like Quality, Safety, Compliance, and Troubleshooting
By leveraging AI models, secure access controls, and an intuitive user interface, this solution empowered Nestlé employees to access relevant information quickly, ensuring they could make data-driven decisions and improve overall operational efficiency