Agentic AI course in Delhi: build AI agents that do real work

AI agents are only useful if they are reliable. You learn to build assistants that answer from a business’s own data, use tools safely and know when to hand over to a human.

Apply for 5 Jan ↗Download brochure ↓

LLM APIsPrompt designRetrieval (RAG)Tool use & agentic workflowsGuardrailsEvaluationAI-assisted engineeringPrivacy & safetyLLM APIsPrompt designRetrieval (RAG)Tool use & agentic workflowsGuardrailsEvaluationAI-assisted engineeringPrivacy & safety

Tool use & agentic workflowsGuardrailsEvaluationAI-assisted engineeringPrivacy & safetyLLM APIsPrompt designRetrieval (RAG)Tool use & agentic workflowsGuardrailsEvaluationAI-assisted engineeringPrivacy & safetyLLM APIsPrompt designRetrieval (RAG)

AI-assisted engineeringPrivacy & safetyLLM APIsPrompt designRetrieval (RAG)Tool use & agentic workflowsGuardrailsEvaluationAI-assisted engineeringPrivacy & safetyLLM APIsPrompt designRetrieval (RAG)Tool use & agentic workflowsGuardrailsEvaluation

Duration4 weeks

Dedicated modules, inside 2 programmes

FormatOffline

In our Shakarpur studio near Nirman Vihar Metro, on real briefs

Learners at work in the Miz Academy studio

learners per batch,
one mentor at the table

ScheduleMon–Fri

10am–5pm · 6 hours a day

Class

Studio practice

Next batch starts5 January 2027Admissions open

Course outline

TEC 202: Agentic AI & AI agents

Companies want AI that saves real time. Builders who can make it dependable, not just impressive in a demo, are in short supply.

What you'll learn

Agentic AI & AI agents: the full skill set

Eight units, each practised on a real or demonstration brief.

LLM APIs

Calling models, handling costs and limits.

Prompt design

System prompts, examples and structured output.

Retrieval (RAG)

Grounding answers in a business’s documents.

Tool use & agentic workflows

Agents that plan steps, call APIs and take actions safely.

Guardrails

Limits, checks and human hand-off.

Evaluation

Testing an assistant before customers use it.

AI-assisted engineering

Using AI to code faster with review habits.

Privacy & safety

What data an AI system should never see.

How you learn it

Learn in the morning, do it by evening

Two classes from 10:00, then four hours of supervised practice. Agentic AI & AI agents is practised on real or demonstration briefs, never just read about.

Built on the web development foundation of the Technology track
Deepest in the 30-week Advanced programme (two full AI agent weeks)
Every assistant is tested by a mentor before demo day

01
10:00 – 12:00
Concept class
The idea behind each agentic ai & ai agents topic, tied to a business problem.
02
13:00 – 15:00
Supervised practice
You do the work on tools, with a mentor in the room.
03
15:00 – 16:00
Mentor review
Your work is marked on a rubric with written notes. You revise.
04
Friday
Present & add to portfolio
Every Friday you present, and the final version joins your portfolio.

Where it's taught

Which weeks cover Agentic AI & AI agents

This skill is part of a full programme, so you also learn everything around it that a working team needs.

Technology · 16 weeks · Job-ready track

AI Technology Strategist

Week 05AI-Assisted CodingLearn: Prompting for code, review, debugging, verificationBuild: AI-assisted feature + review log

See the programme Brochure ↓

Technology · 30 weeks · Career-launch track

AI Technology Strategist Advanced

Week 06AI-Assisted Engineering ILearn: Prompting, code review, debuggingBuild: AI-assisted feature
Week 07AI-Assisted Engineering IILearn: Testing AI code, security pitfallsBuild: Review log + tests
Week 19AI Agents & LLM Apps ILearn: LLM APIs, prompts, tools/function calling, agentic AI patternsBuild: AI assistant v1
Week 20AI Agents & LLM Apps IILearn: RAG basics, guardrails, evaluation, multi-step agentic workflowsBuild: Business AI assistant

See the programme Brochure ↓

What you build

Projects for your portfolio

Business FAQ assistant

Answers from a company’s own documents.

Tool-using agent

An agent that books, looks up or updates records.

Evaluation report

How you tested it and where it fails.

Tools you’ll use

LLM APIsClaude / ChatGPT / Geminin8nNode.jsSupabase / PostgresVercel

Roles this prepares you for

AI Solutions Associate
Junior AI Developer
Automation Developer

Depending on your skills and employer needs. Placement support is not a job guarantee.

Free PDF · 1 MB

Download the AI Technology Strategist brochure

See exactly where Agentic AI & AI agents fits in the week-by-week plan, plus fees and how admission works.

Download brochure ↓Or visit the studio

FAQs

Questions about learning Agentic AI & AI agents

Do I need to know coding first?

The Technology programmes teach coding from the start, then build up to AI agents. The deepest AI work is in the 30-week Advanced programme.

Can I join only for AI agents?

Not yet. AI agents is taught inside our full programmes (AI Technology Strategist, 16 or 30 weeks), because it works best alongside the other skills a working team needs. You learn it on real briefs, reviewed by a mentor.

Is it online or offline?

Offline. Classes and supervised practice happen at our studio in Shakarpur, East Delhi, Monday to Friday, 10am–5pm, near Nirman Vihar and Laxmi Nagar metro.

Related skills

Ten seats per track. Starting 5 January 2027.

Visit the studio, meet a mentor and attend a demo class before you decide.

Apply for the 5 Jan cohort ↗Book a studio visit