Description:

The digital age needs tools that shall help in simplifying our interaction with modern-day technology. An AI desktop assistant can bring significant gains in productivity by automating many mundane tasks. This workshop is designed to arm participants with the know-how to build a simple yet powerful AI desktop assistant using Python. We will learn how to build an Agent Framework that is used to interface a Language Model and allow the model to execute specific commands, such as taking a screenshot and making changes in our recycle bin. By the end of the workshop, attendees will have a working AI assistant and a solid foundation in extending its capabilities.

Workshop Outline:

1. Introduction to AI Desktop Assistants and Agent Frameworks (60 mins)

• Overview of AI desktop assistants and their practical applications
• Introduction of agent frameworks and their role in AI assistants
• How an agent can interact with language models to execute commands
• Live coding session: Develop the foundational agent framework for interpreting and processing user inputs

2. Developing a Screenshot Tool (45 mins)

• Review of the requirements for a 'Screenshot' tool
• Development of a Python tool that captures screenshots
• Integration of the tool with the agent framework for the seamless execution of the developed tool by using prompts
• Live demonstration: Implement a command like \"take a screenshot and save it as test\"

3. Developing a Recycle Bin Management Tool (45 mins)

• Review of the requirements for a 'Management of Recycle Bin' tool
• End-to-end coding of the tool for cleaning the recycle bin, restoring the deleted file, and recovering a specific file
• Integration of the tool with the agent framework for prompt-based operation
• Practical session: Execute commands to manage the recycle bin using the AI assistant

4. Extending and Enriching the AI Assistant (20 mins)

• Discuss possible enrichment/extensions
• Discuss the significance of the Chain of Thought framework in the whole AI implementation process.
• Basic implementation of Chain of Thought to capture the state and enhance the reasoning ability of the AI
• QnA for answering the questions and doubts from the audience

Key Takeaways:

• Develop an in-depth understanding of how to build an agent framework.
• Hands-on experience in building practical tools, such as a screenshot tool and a recycle bin management tool.
• Knowledge of the way to integrate these tools with an AI Assistant framework to provide prompt-based functionality.
• Insights to use Chain of Thought framework to expand the AI capabilities.

Conclusion:

In this workshop, we bridge the gap between theoretical knowledge and practical skills. By the end, you will learn to design your own AI desktop assistant to do whatever you need or want, increasing your productivity and programming abilities.

Prerequisites:

Knowledge of basic Python programming, OOPS principles

Video URL:

https://youtu.be/Y_zPDYsE7zo

Speaker Info:

I am a skilled Python developer with a strong background in machine learning and AI, currently working as a Junior AI Engineer at PathOr Platforms. My technical expertise includes Python, SQL and various AI technologies like OpenAI and Langchain. I have led multiple projects, such as developing AI-driven chatbots and predictive models, demonstrating my ability to integrate advanced machine learning techniques into practical applications. I actively contribute to open-source projects and have a keen interest in emerging technologies and data science. I'm currently a final year B.Tech CS student at Government Model Engineering College, Kochi, Kerala. I'm very active in the community, serving as the founder and chairperson for NSDC MEC, datascience chapter in our College. I share most of my experiences in Linkedin and twitter. You can checkout more about me here.

Speaker Links:

https://www.linkedin.com/pulse/build-your-own-chatbot-langchain-ollama-karthik-g-kumar--aejoc/

https://youtu.be/F8WSxp4zskQ?si=Y1Yd6tLbn8PKZr9M

https://github.com/karthikgkumar/Python-Summer-Internship

https://karthikgkumar.hashnode.dev/polynomial-representation-part-1