Generative AI foundations
Understand tokens, context windows, embeddings and the strengths and limits of language models. Set up Python, notebooks, Git and a first API call.
What you build: Compare model responses against a clear user task.
LIVE TRAINING · MENTOR-LED PROJECTS
Generative AI Course
Turn models into useful applications. Learn the technologies. Build the application. Explain the decisions.
LIVE ONLINE · BUILD WITH A MENTOR
What will you learn in the Generative AI course?
The BAIC Generative AI course is a live online programme for building applications with large language models. You learn prompt design, model APIs, structured outputs, embeddings, vector search and retrieval-augmented generation (RAG), then test and deploy an application that answers from source material.
THE STACK, EXPLAINED
Technologies you will work with
Understand what each tool does, where it fits and how to explain your choices.
- 01Python, notebooks & Git
- Explore model behaviour, build repeatable experiments and version application code.
- 02LLM APIs & prompt templates
- Call models, structure instructions, manage context and return validated JSON.
- 03LangChain
- Connect prompts, model calls, document processing and retrieval into an application.
- 04Embeddings & Chroma
- Represent text for similarity search and store document vectors with metadata.
- 05RAG & retrieval evaluation
- Connect answers to source passages, compare chunking strategies and measure retrieval quality.
- 06Hugging Face Transformers
- Explore open-model interfaces and understand model selection and fine-tuning concepts.
- 07Streamlit & FastAPI
- Build an interface for the document assistant and expose a validated application API.
- 08Docker, pytest & evaluation datasets
- Package the application and check behaviour, response quality, failures, latency and cost.
YOUR LEARNING SEQUENCE
From foundations to a working project.
Prompt engineering & structured outputs
Design instructions and examples, build reusable prompt templates and validate JSON outputs. Handle incomplete responses and input constraints.
What you build: Build a structured information-extraction workflow.
Embeddings, documents & vector search
Load sample documents, split them into useful chunks and attach metadata. Create embeddings, store vectors in Chroma and inspect retrieved passages.
What you build: Build a searchable document collection.
Retrieval-augmented generation
Connect retrieval to model generation with LangChain. Return source references, manage context and respond appropriately when the documents cannot answer a question.
What you build: Build a document assistant with grounded answers.
Evaluation, model choices & fine-tuning concepts
Create an evaluation dataset, inspect retrieval failures and compare prompts and models. Explore Hugging Face interfaces and when retrieval, prompting or fine-tuning fits a task.
What you build: Produce a quality report covering supported and unsupported questions.
Application delivery & capstone
Build a Streamlit interface and FastAPI endpoint, protect configuration, package with Docker and test failures. Measure latency and cost and explain your design.
What you build: Demonstrate and document the completed document assistant.
THE WORK YOU CAN SHOW
A document assistant that answers with evidence
Build an assistant over a controlled collection of documents. Ingest and chunk the material, retrieve relevant passages and return an answer with source references. Test unanswerable questions, compare retrieval choices and report quality, latency and cost.
What goes into your portfolio
- A document ingestion and retrieval pipeline
- A usable application that returns source-backed answers
- An evaluation set with answer-quality and retrieval checks
- A versioned API, Docker image and project walkthrough
BEFORE YOU CHOOSE
Generative AI course questions
What technologies are taught in the Generative AI course?
The learning stack covers Python, LLM APIs, prompt templates, structured outputs, LangChain, embeddings, Chroma, RAG, Hugging Face Transformers, Streamlit, FastAPI, Docker, Git and pytest. Each technology is connected to a practical application task.
Does the course cover RAG?
Yes. You practise document loading, chunking, embeddings, vector search, retrieval, grounded generation, source references and evaluation through a document-assistant project.
Is this a no-code course?
This is an engineering course that uses Python and APIs. If you need programming foundations, discuss the Python for AI Engineering programme before enrolling.
Should I choose Generative AI or Agentic AI?
Choose Generative AI to build grounded model-powered applications and strengthen retrieval and evaluation foundations. Choose Agentic AI when you are ready to build workflows that call tools, maintain state and require human control.
Does the course include fine-tuning?
The syllabus covers fine-tuning concepts and how to compare that approach with prompting and RAG. The main practical project focuses on a grounded application rather than training a foundation model from scratch.
What is the Generative AI course fee?
The Generative AI course fee is ₹14,999, including applicable taxes. Ask a course advisor for payment details and the next batch dates.
Your next step starts with a conversation.
Tell us your experience and what you want to build. Ask about the syllabus, batch dates and fees.
Compare the Agentic AI course or explore AI Applied Engineering.
Generative AI course questions
How long is the Generative AI course?
The course includes 45 live hours, plus hands-on exercises. Confirm the weekly timetable with your course advisor.
Your lead mentor
Gaurav S — Lead FDE, working with multinational clients (15 years of IT experience; previously at HCL, Accenture, HP and Wipro)
Ask about project feedback, the teaching approach and the support available for your course.
View Gaurav’s profile →