Jarvis AI Agent

Back to Portfolio

Jarvis.v1 is an Ironman-inspired AI assistant designed to perform tasks beyond simple conversation. The initial goal was to develop an AI agent that could communicate like Jarvis from the Ironman series, offering a more interactive and functional experience than a typical language model. The project was inspired by ConceptBytes and aimed to integrate various features such as weather updates, time information, music control, image searches, and web browsing.

Features
  • Spotify Controls: Allows control over Spotify playback, including play, pause, skip, and volume adjustment (requires Spotify to be open and playing)
  • Image Search: Performs image searches and saves the results to a designated folder
  • Web Search: Conducts web searches and provides relevant information
  • Youtube Search: Conducts searches on Youtube for requested video
  • Streaming Services Access: Opens streaming service websites for quick access
  • Time Information: Gives the current time upon request
  • Application Launching: Can open specified applications installed on the system
Tools and Technologies Used:
  • Python
  • OpenAI API
  • RealtimeSTT
  • Whisper for TTS
  • iCrawler
  • Asyncio
Jarvis.v2 on the works
VSCode Integration
  • Project Creation: Ability to create new project folders within a specified directory
  • File Creation: Can create new files within project folders to streamline the coding process
  • Code Generation: Generates code based on user requests by querying the OpenAI API and writes the code into specified files (e.g., "Jarvis, create a Python calculator program")
  • Terminal Navigation: Enhanced ability to navigate directories and recognize system paths to improve control over VSCode
ToDo Application
  • Voice-Controlled Task Management: Users can add and manage tasks through voice commands
  • Due Dates and Logging: Assigns due dates to tasks and logs the date and time when tasks are added
  • Detailed Task View: Clicking on a task opens a window displaying details such as task description, logged time, due date, and current stage
Google Calendar Integration
  • Calendar Management: Ability to create, delete, update, and retrieve events in Google Calendar using the Google Calendar API
  • Natural Language Scheduling: Users can schedule events using natural language (e.g., "Schedule a meeting tomorrow at 3 PM")
  • Date Recognition: Understands references like "today," "tomorrow," or specific days of the week for event scheduling
View on GitHub
Web Search, Image Search
Streaming Service feature