AskDex is a secure, Retrieval-Augmented Generation (RAG) chatbot designed to help employees interact with internal policy documents—like privacy statements, health insurance policies, and compliance protocols—without the worry of data leakage or AI hallucinations.
The core goal of AskDex is to provide an intuitive and trustworthy interface for users to ask questions about uploaded documents. It ensures responses are grounded only in the uploaded content, with no generalized or hallucinated feedback — making it ideal for enterprise use cases.
- React
- TailwindCSS
- JavaScript
- Java
- Spring Boot
- OpenAI GPT APIs
- 📄 Document Upload: Accepts plain text documents (e.g.,
.txtformat) containing policy content. - 🔍 RAG-Powered QA: Users ask questions; answers are sourced directly from the uploaded document.
- 🛡️ Privacy-Focused: Designed for on-premise deployment to avoid external API data leaks.
- ✅ Zero Hallucination Policy: No guessing — if it's not in the document, it's not in the answer.
- Enterprise IT & HR Departments: Quick access to company policy documents.
- Insurance Providers: Instant clarification on benefits, coverage, exclusions.
- Privacy & Compliance Teams: Easier navigation of data protection and legal policies.
git clone https://github.com/your-username/askdex.git
cd askdexcd frontend
npm install
npm startcd backend
./mvnw spring-boot:runspring.ai.openai.api-key=YOUR_OPENAI_API_KEY
If you’re interested in collaborating or have feedback on this project, feel free to connect with me on LinkedIn.