Skip to content
MindXcel

Domain 09

Generative AI & LLMs

How large language models actually behave, and how to build on them without the result being a demo that falls over the moment a real user touches it.

Emerging-tech trackPython fundamentals are expected2 programs available
Request counsellingAll domains

Learning objectives

What you will be able to do.

  • Explain what an LLM is doing well enough to predict its failures
  • Design and iterate prompts systematically rather than by feel
  • Ground responses in your own data using retrieval
  • Build a retrieval-augmented application end to end
  • Evaluate generated output beyond reading a few samples

Who this is for

  • Learners with programming experience moving into AI engineering
  • Developers who want to add AI capability to real products

Prerequisites: Python fundamentals are expected

Tools and technologies

  • Python
  • LLM APIs
  • Embedding models
  • Vector databases
  • Jupyter
  • Git

Roles this prepares you for

  • AI Engineer
  • LLM Application Developer
  • AI Product Engineer

Curriculum

5 modules, in the order they build on each other.

Each module ends in something you have made work, not just something you have watched.

  1. 01

    LLM foundations

    • Tokens, context windows and cost
    • How generation works
    • Model selection and trade-offs
    • Temperature and sampling
  2. 02

    Prompt engineering

    • Instruction design
    • Few-shot and structured output
    • Iterating against a test set
    • Prompt injection and safety
  3. 03

    Embeddings and retrieval

    • What embeddings represent
    • Vector stores and similarity search
    • Chunking strategies
    • Hybrid retrieval
  4. 04

    RAG architecture

    • The retrieval pipeline
    • Grounding and citation
    • Handling the no-answer case
    • Latency and caching
  5. 05

    Evaluation

    • Building an evaluation set
    • Automated and human evaluation
    • Regression-testing prompts
    • Monitoring in production

Practical work

What you could build in this domain.

Project briefs are agreed with your mentor at the start, so the work suits the level you are actually at.

  • 01

    A document Q&A assistant with source citations

  • 02

    A structured extraction pipeline over unstructured text

  • 03

    An evaluation harness comparing two prompting strategies

Programs

How to study Generative AI & LLMs.

The same domain, in the format that fits your situation.

All programs
  • Generative AI & LLMsIntermediateInternship included

    Generative AI Internship

    Build an LLM-powered application — prompting, retrieval and evaluation — on a problem you pick.

    • Prompt design and systematic iteration
    • Embeddings and retrieval over your own documents
    • Grounding answers and citing sources
    • Evaluating output quality beyond eyeballing it

    You finish with: A working retrieval-augmented application over a document set of your choice.

    Program details
  • Generative AI & LLMsIntermediate to Advanced

    Generative AI & LLM Applications

    How large language models actually behave, and how to build applications on them that stay grounded and measurable.

    • LLM concepts, context windows and model selection
    • Prompt engineering as an iterative discipline
    • Embeddings, vector stores and retrieval
    • RAG architecture and grounding strategies
    • Evaluating generated output systematically

    You finish with: A retrieval-augmented application with an evaluation harness.

    Program details

Keep looking

Other domains.

Learn Smart. Build Practical. Grow with MindXcel.

Learn the technologies shaping tomorrow — AI, ML, Data Science, Generative AI, LLMs and AI Agents.

Request counsellingCompare programs