Overview

Learning Objectives

The 400-level section will cover the growing ecosystem of generative AI tools, technologies and platforms from a developer perspective.

• Fundamental Terms & Concepts for Generative AI
• Core Developer Tools, Technologies & Platforms
• Popular Models, Model Providers & Applications
• Cookbook: Recipes for frequently-done tasks
• Solution: End-to-End workflow for one signature app

01 | Fundamental Concepts

Generative AI has a number of core concepts we need to know more about - like prompt engineering, model fine-tuning, responsible AI, evaluation, LLM Ops etc. Use this section to provide links to relevant papers, tools and tutorials that teach fundamentals.

02 | Developer Tools

Building generative AI applications requires tools for building apps, evaluating models, deploying solutions, and orchestrating end-to-end workflows. Use this section to provides links to relevant libraries, frameworks and platforms that generative AI developers should know.

03 | Provider Ecosystem

Generative AI applications are driven by language models - both Large Language Models (pre-trained on massive datasets) and Small Language Models (fine-tuned for specialized domains), and variants based on needs e.g., Large Multi-Modal Models (LMMs). Use this section to track the growing model providers ecosystem (and relevant models) gaining mindshare with developers. Some examples to start with:

Provider	Resources
OpenAI	Models : Embeddings · Moderation · GPT · DALL-E · Whisper · TTS · Sora ▫️ Apps : ChatGPT ▫️ Responsible AI : Safety
Azure AI	Models : Catalog · Benchmarks · Prompts ▫️ Apps : Contoso Chat ▫️ Responsible AI : Tools
Hugging Face	Models : Model Hub · Hugging Chat · Courses · Cookbook ▫️ Apps : Society of Ethics ▫️ Responsible AI : Principles
Google AI	Models : Gemini · Palm · Imagen · MedLM · Codey · Chirp · Gemma (open models) ▫️ Apps : Examples ▫️ Responsible AI : Principles

04 | Developer Cookbook

This section and the next are focused specifically on developers and capturing tutorials or “recipes” that cover common tasks or workflows in generative AI applications. These should ideally be code-first (snippets) with documentation (what, why) in notebooks.

05 | End-to-End Samples

This section captures real-world applications samples that show end-to-end developer workflows from ideation to operationalization, potentially with different platforms (cloud) and providers (model) to get developers a sense or practical challenges and design patterns for production use.