The Virtual Assistant JARVIS
JARVIS is an AI-driven virtual assistant aimed at streamlining daily business processes and enhancing routine tasks. Developed on the n8n automation platform, it allows for adaptable workflow management and smooth integration with various services and APIs. The project aimed to create a smart assistant that could handle communication, schedule tasks, gather information, and offer context-aware responses based on user interactions.
**Key Objectives:**
- Automate email handling and regular communication
- Organize meetings, reminders, and calendar events
- Retrieve information from external knowledge sources
- Provide personalized responses utilizing contextual memory
**System Architecture:**
JARVIS functions through a set of specialized AI agents, each assigned different responsibilities:
- `emailAgent`: Manages emails by reading, composing, responding, and organizing inbox messages
- `calendarAgent`: Oversees meeting arrangements, checks availability, and sets reminders
- `JarvisMemory`: Maintains contextual information and user preferences
**Core Capabilities:**
- Intelligent analysis of queries to ascertain user intent and direct tasks to the suitable agent
- Email automation that minimizes the time spent on repetitive communication
- Calendar management for scheduling and reminder tasks
- Retrieval of external data from sources like Perplexity, SerpAPI, and Wikipedia
- A contextual memory system grounded in a vector database, which facilitates semantic search for retrieving relevant information based on meaning rather than just keywords, while also preserving context throughout interactions.
**Technologies Used:** n8n, Telegram API, Gmail API, Google Calendar API, Perplexity API, SerpAPI, Wikipedia API, Vector Database.
**Results:**
- Automated email and scheduling operations
- Quicker access to relevant information
- Context-aware responses leveraging stored user data
- Scalable AI automation built on the n8n platform.