Generative AI Apps with ChatGPT, LangChain & Hugging Face

Explore generative AI models such as LLMs, GANs, and diffusion, and learn how they create text, images, code, and other content with real-world applications in education, software development, and marketing.

Learn the four pillars of generative ai apps by using OpenAI, LangChain, Hugging Face, and Pinecone. Build real-world multimodal apps through api integration, smart workflows, and semantic search across data.

Master prompt engineering to guide llms like GPT, Claude, and Gemini with clear task, context, persona, format, and training; apply best practices for precise, useful, and structured AI outputs.

Compare api-based models and local models to decide when to use each; APIs offer speed, simplicity, and quick prototyping, while local models provide offline access, privacy, and full control.

Explore the OpenAI playground as a developer toolkit to tailor prompts and model behavior with temperature, top P, and max tokens, demonstrated with a GPT-4.1 mini live demo.

Explore the core building blocks of generative AI apps, including tokens, prompts, completion models and chat models, and memory in multi-turn conversations, with hands-on OpenAI playground demos.

learn to build custom gpt assistants with openai's assistant api using low code, define roles, enable tools like code interpreter and file uploads, and read documents.

Explore ToolAgent and ReactAgent in real-world scenarios, building a weather tool agent to demonstrate practical integrations within generative ai apps using ChatGPT, LangChain, and Hugging Face.

Explore LangGraph to build stateful, multi-step agents with graph-based workflows that incorporate memory and branching logic, enabling search and summarize flows using LangChain and OpenAI models.

Explore Hugging Face's open source hub to plug pre-trained NLP models into apps with a few lines of code, enabling translation, summarization, and emotion detection across languages.

Explore open source models with the Transformers library from Hugging Face to build text generation apps using GPT-2, via a simple five-line code setup and pipeline.

Build a text summarizer or classifier using Hugging Face transformers in Python. Use a Distilbert model in a simple pipeline to summarize text or analyze sentiment in Google Colab.

Explore LoRA fine-tuning, a lightweight low-rank adaptation that injects small A and B matrices to update only the trainable components, keeping original weights frozen for fast, cheaper model adaptation.

Learn how memory in LLMs preserves context across conversations via buffer, summary, entity, and vector store memories, and build a buffer-memory chatbot in Colab with LangChain and OpenAI.

Discover how embeddings power retrieval-augmented generation by turning text into semantic vectors, enabling chunk-based searching, cosine similarity, and chat with custom PDFs via vector stores.

Learn how embeddings power semantic search by storing and querying high-dimensional vectors in vector databases like Chroma and Pinecone, enabling fast similarity retrieval with LLMs.

Compare pinecone, chroma db, and faiss to choose the right vector database for fast, scalable semantic search with embeddings. Understand deployment, persistence, filtering, and cost between offline and cloud use.

Learn to connect vector databases with LangChain to perform semantic search on your data and retrieve relevant chunks to feed to the model for contextual answers.

Build a website retrieval chat app that scrapes pages, creates vector embeddings, and answers questions via semantic search with an llm. Leverage LangChain and OpenAI embeddings for retrieval-augmented generation.

Learn to build interactive web user interfaces for ML apps using Streamlit and Gradio, turning code into dashboards and quick prototypes, with Ngrok sharing.

Build a live chatbot with a flask backend and a custom HTML, CSS, and JavaScript frontend, exchanging messages via fetch in a docker-powered hugging face spaces.