Technology Training • AI & Generative AI
AI & Generative AI Training for Institutions & Organizations
Practical training in building with large language models: prompting them well, grounding them in your own data with retrieval, and wiring them into agents and automated workflows. Learners finish with working applications, not just slides about AI.
What the training covers
Modules are selected and sequenced for each cohort. A short workshop may use one or two; a semester course or corporate program can cover all of them.
Module 1
How large language models work
Enough of the underlying mechanics that learners can reason about cost, limits and failure modes instead of treating the model as magic.
- Tokens, context windows and sampling
- Model families and choosing between them
- PyTorch foundations for neural networks
Module 2
Prompt engineering
Writing prompts as reviewable, testable artefacts: structure, examples, output formats and how to tell whether a change made things better.
- Instruction and few-shot prompting
- Structured outputs
- Evaluating prompt changes
Module 3
Retrieval-Augmented Generation (RAG)
Grounding answers in a document collection so the model cites real sources instead of guessing. Learners build a retrieval pipeline end to end.
- Chunking and embeddings
- Vector search
- LangChain and LlamaIndex pipelines
Module 4
AI agents and autonomous workflows
Letting a model call tools and take multi-step actions, with the guardrails that make it safe to run unattended.
- Tool calling
- Multi-step agent loops
- Workflow automation and human review points
Module 5
Fine-tuning and when not to
When adapting a model's weights is worth it compared with better prompts or retrieval, and what a small fine-tuning run involves.
- Fine-tuning workflow
- Data preparation
- Comparing fine-tuning, RAG and prompting
Who it is for
Colleges & universities
Third- and final-year CSE, IT and AI/DS students who already program in Python, and faculty who want to add generative AI projects or electives to their courses.
Organizations & engineering teams
Software teams adding LLM features to products, and engineers asked to build internal assistants, document search or workflow automation.
Prerequisites and lab setup
What learners should know beforehand, and what your lab or IT team needs to have ready. We confirm the final checklist with you before the program starts.
Learner prerequisites
- Comfortable writing Python functions and working with lists and dictionaries
- Basic understanding of HTTP APIs and JSON
- No prior machine-learning course is required for the LLM and RAG modules; the PyTorch module assumes basic linear algebra
Lab and infrastructure
- One laptop per learner with Python installed and permission to install packages
- Reliable internet access for model APIs
- Model API access: we agree in the proposal whether to use your organization's keys, free tiers or locally run open models
- A GPU is not required for the core modules
How it is delivered
The same ai & generative ai content can be delivered in different formats, on-site or live online.
Technical workshops
A focused workshop works well for prompt engineering and a first RAG application built in one sitting.
Semester training
A semester course can take students from LLM basics through RAG to an agent-based capstone project.
Faculty development (FDP)
Faculty sessions concentrate on designing generative AI assignments and lab exercises that hold up to academic evaluation.
Corporate training
Corporate programs can be built around your use case, such as search over your own documents, with your data-handling rules in place.
AI & Generative AI Training FAQs
Will students build something, or is this only theory?
Every module ends in working code. Typical lab outputs are a prompt test suite, a document question-answering app built with retrieval, and a small tool-using agent. The exact projects are agreed in the proposal.
Do we need paid API accounts or GPUs?
GPUs are not needed for the core modules. For model access we plan with you before the program starts: your own API keys, free tiers, or open models run locally, depending on budget and policy.
Can the content be aligned with our AI/DS syllabus?
Yes. Share your syllabus or elective outline and we map modules to it, so sessions reinforce what is assessed rather than running in parallel to it.
How do you handle our company data in RAG exercises?
We can run exercises on public sample documents or on data you approve. If your policy restricts sending data to external model APIs, we plan the labs around locally run models instead.
Is this suitable for non-programmers?
This program is designed for people who write code. For non-technical staff we can discuss a separate, shorter session on using AI tools safely, scoped in the proposal.
Planning a broader program? See all technology training programs.
Plan AI & Generative AI training for your cohort
Share your cohort, current skill level, timeline and lab setup. We will prepare a tailored syllabus and proposal.