Building Aaron the Intern, a Generative AI Challenge Lab 👩‍💻👨‍💻

Imagine you are the newest engineer at Google Cloud. Your task is helping your GCP customers make their tech work on GCP. You surely would love an intern to do the tedious technical research work for you...

But have you ever tried getting budget for something like that..?! Well, great that you live in the age of AI. So why not just building that intern yourself? Let's call "it": "Aaron the intern".

This lab is designed to guide you through GCP Generative AI, Langchain, and building real-world applications. We'll be tackling a series of challenges, each building upon the last, to transform a basic Next.js application into a powerful AI-driven tool.

Your end result will be an interactive, research agent that can take actions on a flexible knowledge base that you define.

Getting Started with GCP 🚀

Before you embark on your Gen AI adventure, you'll need to set up your Google Cloud Platform (GCP) environment.

1. Create or Select a GCP Project:

Go to the GCP Console, login with your credentials and select the project you have in your sandbox environment.

2. Your IDE and Shell in the Cloud:

• Open your Cloud Shell
• Open youd Cloud Shell Editor
• This is your easiest IDE and Shell environment to work through this lab.
• If you want to work on your local instead, feel free to so and set up auth yourself.

4. Set up GCP Project and Application Default Credentials (ADC):

Run the following commands to authenticate yourself and set your project as default in the console:

gcloud auth application-default login
gcloud config set project <YOUR_PROJECT_ID>

Fork the workshop Content 🚀

1. Fork the GitHub Repository:

• On GitHub create a fork of this repository.
• Make sure to include all branches

2. Clone the fork into your editing environment:

• Back in the Cloud Shell run git clone <url of your forked repo>.

Set Environment Variables:

1. Env Vars in .env.local:

• Create a new file (in your project root) called .env.local based on .env.template
• You will find you Gemini Api Key in AI Studio. You might need to select "Show projects in my organization"
• Fill out the project ID from your Cloud Console

2. Env Vars in terraform.tfvars:

• Find the file terraform/terraform.tfvars in your project root
• Fill out the project ID just as you did in .env.local

Deploy the required Infrastructure:

Deploy the required GCP Infrastructure by running the terraform script as follows:

cd terraform
terraform init
terraform plan
terraform apply

Running the App

After completing the steps above you are able to run the app and access it on localhost with:

npm install
npm run dev

Optionally, you can deploy the web app to cloud run by running make all in root.

Congratulations! You've completed the basic GCP setup. You are now ready to explore the application and start the challenges!

Workshop Mechaniccs

We've organized the challenges using GitHub Branches to keep things clear and manageable as you progress through the challenges. Also the use of branches makes each challenge independent from the others. You are free to tackle the individual challenges in any order.

• main Branch (Your Starting Point): This is where you are right now! It contains the base application (and full solution!) – You can explore the initial setup, run the app, and get familiar with the codebase. This branch will remain untouched as you work through the challenges. You can always come back or compare to main for the challenge solution.

• 001-llm-data-mining, 002-build-knowledgebase, 003-simple-rag-qa, 004-advanced-rag, 005-llm-eval-and-tracing Branches: Each of these branches represents a challenge focussing on different learnings. You'll switch to these branches to work on each challenge individually. Each challenge branch works independent from the other branches.

• Each challenge branch contains a challenge specific instruction Markdown file. For challenge 1 this is 001-llm-data-mining.md. Follow these instructions to find out what to do in each challenge.

To start a challenge, you'll simply switch to its corresponding branch using Git:

git fetch origin
git checkout 001-llm-data-mining # For Challenge 1
git checkout 002-build-knowledgebase # For Challenge 2, and so on...

Don't worry about messing things up! Because each challenge is in its own branch, you can experiment freely. If you ever want to start fresh, you can always return to the main branch or reset your challenge branch back to origin (as long as you don't push your changes).

Your Missions 🏋️‍♀️

Here is an overview of the challenges and the learnings they contain:

Challenge 1: LLM Powered Data Mining 🕵️‍♀️

• Mission: Bring your text prompting to the next level.
• Learnings:
  • Prompt Engineering for Extraction: Crafting effective prompts to instruct the LLM to perform realiable entity extraction.
• Key Technologies: Next.js, Gemini API, Prompt Engineering.

Challenge 2: Create Knowledge Base 📚

• Mission: Implement a custom indexing pipeline on GCP.
• Learnings:
  • Chunking Strategies: Exploring different ways to divide text into chunks for optimal retrieval.
  • Embedding Creation: Understanding how text embeddings are generated and their role in semantic search.
  • Embedding Storage: Learning about vector databases (in this case, Firestore is used for simplicity) for efficient storage and retrieval of embeddings.
• Key Technologies: Next.js, Firestore, Gemini API for embeddings, Chunking algorithms.

Challenge 3: Simple RAG 🤖

• Mission: Build a Basic RAG Q&A pipeline.
• Learnings:
  • Simple RAG Pipeline: Understanding the core components of RAG: Retrieval (searching the knowledge base) and Generation (using an LLM to create answers).
  • Semantic Search: Using vector search to find relevant documents based on meaning, not just keywords. The VectorSearchUtil is your search engine.
  • Prompt Engineering for Summarization: Crafting prompts to guide the LLM to generate concise and accurate summaries from retrieved documents.
• Key Technologies: Next.js, Firestore, Gemini API, Langchain, Vector Search, Prompt Engineering.

Challenge 4: Advanced RAG 📧

• Mission: Build an email generation pipeline based on an advanced RAG invocation chain.
• Learnings:
  • Prompt Chaining with Langchain: Using Langchain's RunnableSequence to create complex, multi-step workflows.
  • Advanced RAG Techniques: Building a more sophisticated RAG system that involves task extraction, research, and email generation.
  • Prompt Engineering for Email Generation: Designing prompts to create emails that are not only informative but also professional and customer-friendly.
• Key Technologies: Next.js, Firestore, Gemini API, Langchain, Vector Search, Prompt Chaining, Advanced Prompt Engineering.

Challenge 5: Tracing & Tracking 🔭

• Mission: Observce and monitor your LLM calls.
• Learnings:
  • Tracing and Tracking: Understanding the importance of observability in LLM applications.
  • Langfuse Integration: Using Langfuse to trace requests, monitor performance, and analyze results.
  • Quality Control & Experimentation: Using tracing data to identify areas for improvement and to evaluate the impact of changes (like prompt modifications).
• Key Technologies: Next.js, Langfuse, Langchain, Observability principles.

Get Ready to Explore! 🚀

Each challenge will have its own detailed README.md file within its branch, providing specific instructions and hints.

So, are you ready to build yoru own "Aaron the intern" and master the power of Generative AI on GCP? Let's get started! Head over to challenge-1 branch to begin your adventure! Good luck, and have fun! 🎉