Veldora is a next-generation AI assistant designed to overcome the limitations of traditional chatbots by combining memory, reasoning, and real-time data integration.
Unlike standard AI systems that rely only on pre-trained knowledge, Veldora uses a hybrid architecture that includes Short-Term Memory (STM) for maintaining conversational context and Long-Term Memory (LTM) powered by embeddings for intelligent recall. This allows the assistant to understand users better over time and provide more personalized, context-aware responses.
To address the challenge of outdated information, Veldora integrates external APIs and a decision-making layer that determines when real-time data is required. This enables the system to fetch and use up-to-date information, making it more reliable for real-world use cases such as current events, dynamic queries, and live problem-solving.
The assistant is also built with multilingual capabilities, allowing it to communicate seamlessly across different languages, making it accessible to a wider audience.
Veldora is not just a chatbot—it is an AI agent that combines knowledge, tools, and memory to deliver accurate, adaptive, and intelligent interactions. The project aims to create a scalable foundation for future AI systems that can think, learn, and respond more effectively in real-world environments.
Built With
- api
- embedding
- express.js
- gemini
- javascript
- mongodb
- node.js
- socket.io
- vector
Log in or sign up for Devpost to join the conversation.