Module 6: Project Tracks and Submission

Note: The last day to submit your project will be conveyed via email

As we approach the conclusion of this bootcamp, we'll apply your acquired knowledge to practical applications.

To guide you, we have selected a range of ideas/tracks (which are optional) but you can use them for ideation if needed. However, we encourage you to go beyond and think of a few ideas by looking at the problems around you, as that's one of the better approaches to problem solving. 🌟

You should also explore the resources listed under prerequisites so the hands-on module is easier for you to finish. If you're done with that as well, you could share your learning journey with us and the world out there; learning in public comes with a dozen advantages anyway.

Now, let's quickly revisit the mandatory requirements for completing the bootcamp.

Criteria for Successfully Complete the Bootcamp

1 – Complete the Quizzes as part of Cohort based learning

  • Ensure you complete the required quizzes: one in the Vector Embeddings module and another in the RAG module, which will be released as per the schedule.

2 – Project Development

  • Task: Develop a real-time or static RAG-based LLM application completely using Pathway LLM App templates or Pathway with Llamaindex / Pathway with Langchain.
  • Publish: Publish your open-source project on your GitHub with a clear README that includes a video demo. Please We emphasize this as it makes it easy for course instructors, developers in the community, or your potential employers to evaluate what you've built. To streamline the process of creating a README, you can use AI tools like ChatGPT or Gemini. Please find this document for "Sample prompt for README Generation".
  • Submission: Submit the project.

3 – Project Guidelines

  • Option to Modify an Existing Project: If building an LLM application from scratch seems daunting, consider modifying an existing one we've seen earlier discussed. Adapt it to create an application with significant business or social value. For inspiration, look at how Avril adopted an existing RAG project to better comprehend the EU AI Act. This being said a direct replica of any published project will not be accepted.
  • Project Requirements:
    • Data Source: Your project should use dynamic (preferred) or static data sources.
    • Open Source: Ensure your project is open source, hosted on GitHub with a clear README.md file and a License file as a best practice. Ref: Adding a License to a Repository / Tutorial for adding MIT License.
    • Documentation: The README.md must include:
      • A demo video link or GIF for a quick overview of your application.
      • A clear description explaining the purpose of your project and how it utilizes Pathway, Langchain, LlamaIndex, Ollama (sample documentation), etc.
      • Instructions for setting up and running the tool.
  • Originality: Your project must be original, not plagiarized, and not a direct replica of any course materials, publicly available projects, or those submitted by peers.
  • Production Readiness: Deploy your project into one of the public clouds. You can refer to our AWS, Azure, and GCP deployment tutorials. Deploying to the cloud helps ensure your project is stable, resilient to failures, and meets modern industry standards.
  • Bonus: If you publish your project as a tutorial on any popular developer publication (e.g. Freecodecamp, Dev.to, GFG, KDNuggets, Towards Data Science, etc.) it becomes significant proof of clear documentation and implementation for the team at Pathway and your future collaborators/employers. However, at times it may take additional efforts (simply copy-pasted Gen AI articles also need refinement) so it's not a mandatory thing. But its importance cannot be overstated.

Encouragement for Innovation 💪

  • While using an existing open source project as a foundation is acceptable, we encourage you to innovate and create something unique. Challenge yourself to develop a project that tests your cognitive abilities and engineering skills.
  • If the idea of creating an LLM application from the ground up (like the one we saw in the Amazon Discounts case) feels overwhelming, you have the option to build upon the existing examples discussed earlier. By tailoring it to meet specific needs, you can construct an application that holds substantial business or social value.

What are additional incentives beyond learning for building a novel application? Let's see 🤩