GenAI for Developers

Overview

This course provides a deep dive into Large Language Models (LLMs) and Generative AI, covering essential concepts, frameworks, and advanced applications. Participants will learn how to build AI-powered applications, optimize workflows, and implement security best practices in AI-driven systems.

Since this a practical, project focused course, we are not going to focus on math part rather our goal is to use AI tech stack to build application/products.

What You Will Learn

• Introduction to LLM and Generative AI – Understanding the fundamentals of LLMs and their capabilities.
• AI Agents and Agentic Workflows – Implementing intelligent, autonomous AI agents.
• Building Basic Chat Applications – Using LangChain to develop AI-driven chatbots.
• Chat Over Large Documents – Leveraging vector stores such as Qdrant DB, PG Vector, and Pinecone for efficient document retrieval.
• Retrieval-Augmented Generation (RAG) – Enhancing AI responses with dynamic information retrieval.
• Context-Aware AI Applications – Developing AI solutions that adapt to different contexts.
• Memory-Aware AI Agents – Utilizing Qdrant DB and Neo4j Graph for persistent AI memory.
• Document-to-Graph DB and Embeddings – Transforming structured and unstructured data into graph-based representations.
• Multi-Modal LLM Applications – Integrating text, images, and other data modalities.
• Security and Guardrails – Implementing self-hosted models like Llama-3 or Gemma to ensure AI safety and compliance.
• AI Agent Orchestration with LangGraph – Managing multiple AI agents and workflows.
• Checkpointing in LangGraph – Ensuring fault tolerance and reproducibility in AI pipelines.
• Human-in-the-Loop Interruptions – Allowing human oversight in AI-driven decisions.
• Tool Binding and API Calling – Enabling AI agents to interact with external tools and services.
• Autonomous vs. Controlled Workflows – Understanding different agent workflow strategies.
• MCP Servers – Deploying and managing AI microservices efficiently.
• Guardrails for AI Models – Implementing prompt filtering, PII detection, and safety mechanisms.
• Model Fine-Tuning – Customizing pre-trained LLMs for specific use cases.
• LLM as a Judge Technique – Evaluating AI-generated responses using AI.
• Perplexity Sonar API – Enhancing AI reliability and accuracy.
• Deployment on AWS – Hosting AI applications on a scalable cloud infrastructure.
• Cypher Query Context Retrieval – Enhancing LLM capabilities with Neo4j Graph DB.

Tech Stack

This course will utilize the following technologies to build, optimize, and deploy AI applications:

• Programming Language: Python
• LLM Models: OpenAI, DeepSeek, Claude
• Frameworks:
  • LangChain – A framework for building AI-powered applications.
    • LangGraph – A tool for structuring and managing AI agent workflows.
    • LangSmit – Enabling efficient AI development and execution.
• Tracing & Monitoring:
  • Langfuse (Docker) – Self-hosted traces for AI applications.
• Memory and Vector Stores:
  • PG Vector – A high-performance vector database.
  • Quadrant DB – A scalable, efficient vector store.
  • Vector Embedding Models – Enhancing AI understanding through embeddings.
• Infrastructure:
  • MCP Server – Managing AI inference and computation.
  • Neo4j Graph DB – Graph-based AI knowledge storage.
  • AWS – Scalable cloud deployment for AI applications.

Learning Outcomes

By the end of this course, participants will gain expertise in:

• Frameworks: Mastering LangChain, LangGraph, and Hugging Face Transformers.
• Databases: Implementing Qdrant, Neo4j, and Pinecone for AI applications.
• Models: Understanding OpenAI, Gemini, Llama-3, and Gemma.
• Infrastructure: Deploying AI solutions using AWS, Docker, LangSmit, and Langfuse.

Hands-On Projects

Participants will apply their knowledge by developing real-world AI projects, including:

AI-Powered Legal Document Assistant – Automating legal document processing and summarization.

AI-Powered Chart Builder with Postgres – Generating interactive data visualizations using AI.

AI-Powered Resume Roasting – Evaluating and improving resumes with AI-driven feedback.

AI-Powered Candidate Search – Enhancing recruitment with intelligent candidate matching.

AI-Powered Website Bot – Enabling AI-driven interactions with website content.

This course provides a hands-on approach, ensuring learners gain practical experience in building and deploying AI applications at scale.

Mentors

@hiteshchoudhary @piyushgarg-dev