No prior coding experience required

Gen AI and Agentic AI Mastery

A 12-week, hands-on programme plus a 1-week capstone project for non-tech professionals to build AI workflows, assistants and agents.

Apply now Explore curriculum
IIT Patna campus at Bihta
  • 12 + 1 weeksCapstone project included
  • Live and OnlineFormat
  • 6–8 hrsPer week
  • 30 toolsHands-on
Why choose this course?

Lots of people can use AI.
Few can build with it.

You don't need to code to build with AI. You need to know how it works and which tools to use.

No coding required

Every week is built around no-code tools and Claude Code, so you build real systems without learning to programme.

Up to date

Covers the newest agent tools, such as Jev, OpenAI's Decisions API and OpenClaw or Dots, taught concept-first so your skills outlast any one vendor.

Production, not just demos

Two full weeks on quality and production: evals, guardrails, tracing and cost control.

Open and closed models

Run open-source models on your laptop or in the cloud (Ollama, LM Studio, Hugging Face) and learn when to pick them over frontier models.

Beyond chat

A dedicated week on image, video and voice, from diffusion models to real-time voice agents.

Who is this for?

Built for non-tech professionals.

No coding background needed. If you use a computer at work, you're ready.

A product manager reviewing work on a tablet

Product managers

Want to prototype AI features, write sharper AI specs and judge vendor claims without waiting on engineering.

An operations professional working on her laptop

Ops teams

Ops, support and business teams automating tickets, reports and hand-offs with agents and n8n workflows.

A founder standing with arms crossed in his office

Founders

Founders and solopreneurs building and shipping an AI product or internal tool without hiring developers.

A young professional smiling at his laptop

Career switchers

Career switchers and early‑career professionals moving into AI roles who need a portfolio, not just a certificate.

Curriculum

Twelve weeks, one build at a time.

Four phases, from your first API call to AI in production, then a final week for your capstone project.

  • APIs & SDKs: how apps talk to AI
  • Webhooks that trigger AI from real events
  • Building your first LLM-powered workflow
  • When AI should decide vs when rules should

Tools you will learn

PostmanFastAPICursor
  • What actually happens inside an LLM
  • Prompt patterns & chain-of-thought
  • Diagnosing & fixing weak outputs
  • Matching the right model to the job

Tools you will learn

OpenAIClaudeOpenRouter
  • Embeddings & vector search, demystified
  • Splitting documents & retrieving what matters
  • RAG patterns and when each fits
  • Going further: graph-based & agentic RAG

Tools you will learn

LangChainPineconeNeo4j

*Please note: "Due to the evolving nature of industry expectations and partner institute feedback, some syllabus aspects may change. Any updates will be communicated during the Inauguration Session(s) or at the start of the relevant module."

Capstone projects

Two real-world builds for your portfolio.

Project 1

Swiggy Customer-Support Agent

Build a multi-agent support system in LangGraph where tool‑equipped agents (triage, refund-handler and escalation-writer) hand off through conditional routing. Test it on 30 simulated tickets, measure resolution accuracy and write up the architecture.

Skills
  • LangGraph
  • Multi-agent design
  • Conditional routing
  • Architecture documentation
OutcomeProduction-grade multi-agent system with eval results
Project 2

Cred Knowledge-Base RAG System

Build a RAG system over public sources (credit-card terms, RBI guidelines and Cred help articles) with LangChain and a vector DB (Chroma). Evaluate retrieval on 30 test queries and report grounding accuracy.

Skills
  • LangChain
  • Chroma/Pinecone
  • Vector embeddings
  • Evaluation
OutcomeWorking RAG system with quantified retrieval accuracy

*Please note: Project deliverables may vary by batch. Brand names are used for learning purposes only; student projects are independent concepts and are not endorsed by the respective brands.

Tools you will learn

The stack teams actually use.

30 tools across models, retrieval, agents, evaluation, deployment and multimodal AI.

  • PostmanAPIs
  • FastAPIAPIs
  • CursorAI IDE
  • OpenAIModels
  • ClaudeModels
  • OpenRouterModel routing
  • LangChainRAG
  • PineconeVector DB
  • Neo4jGraph DB
  • n8nAutomation
  • LangGraphAgents
  • Jev / OpenAI Decisions APIDecisions
  • promptfooEvals
  • TruLensEvals
  • MixpanelAnalytics
  • Claude CodeCoding agent
  • GitHubDev workflow
  • CodexCoding agent
  • OpenClawAssistants
  • Hermes AgentAssistants
  • Claude in ChromeComputer use
  • LangfuseObservability
  • VercelDeployment
  • Hugging FaceOpen models
  • OllamaLocal models
  • LM StudioLocal models
  • GeminiMultimodal
  • RunwayVideo
  • ElevenLabsVoice
  • DeepgramSpeech
Instructors

Learn from people who ship AI.

Practitioners building AI products in industry, not just teaching about them.

Portrait of Deepak Kasera

Deepak Kasera

Microsoft

Senior Software Engineer working on cloud, AI at scale and agentic AI. Previously at Amazon.

LinkedIn
Portrait of Prof. Priyam Chakraborty

Prof. Priyam Chakraborty

IIT Kharagpur

Assistant Professor of Aerospace Engineering. Previously Lead Data Scientist at Happymonk AI Labs.

LinkedIn
Portrait of Mohit Uniyal

Mohit Uniyal

PayPal

Senior ML Engineer who has taught 2,000+ hours of data science and ML. Previously at Walmart Global Tech.

LinkedIn
Portrait of Ashish Prasad

Ashish Prasad

Microsoft

Engineering Manager for AI and ML at Microsoft Turing. Helped bring GPT‑5 into Microsoft 365 Copilot.

LinkedIn

*Instructors may change based on their availability.

Outcomes

What you'll walk away with.

01

Build without code

Ship workflows, agents and apps with n8n and Claude Code, and deploy them with Vercel.

02

Design agents and RAG on your own data

Context engineering, decision models (Jev, OpenAI Decisions API), MCP, graph and agentic RAG, plus computer‑use and always‑on assistants.

03

Make AI reliable and affordable

Test with evals, add guardrails and human approvals, trace runs in Langfuse, and know when an open-source model is the cheaper, safer choice.

04

Graduate with a portfolio

Real-world capstone projects: a Swiggy support agent and a Cred RAG system.

Certificate

Earn a certificate from IIT Patna.

Complete the programme to receive a certificate from Vishlesan i‑Hub, IIT Patna, in collaboration with Masai.

Sample certificate of completion for Gen AI & Agentic AI Mastery from Vishlesan i-Hub, IIT Patna
Admissions

Three steps to get started.

A learner working on her tablet and laptop
Step 01

Submit application

Complete the application form to showcase your motivation and goals.

Step 02

Complete selection and enrolment session

Only shortlisted candidates go through the counselling process.

Step 03

Secure your seat

Learn from India’s top educators and stand out from the crowd.

Programme details

Everything you need to know.

Total fee
₹29,999+ GST
Apply now
Application feeNot included in the total fee.
₹49(Non-refundable)
Registration feePaid upfront.
₹4,000(Non-refundable)
Programme feePay in full, or in 3 monthly EMIs of ₹9,204 + GST (₹31,612 + GST in total with EMI)
₹25,999

The application fee and registration fee are non-refundable. The programme fee follows the Refund & Withdrawal Policy.

FAQ

Questions, answered.

What is the duration and weekly effort of the programme?

The programme runs for 12 weeks.

Is the programme live or recorded?

Live. Every class and doubt session is held live.

What is the mode of delivery for the programme?

Live and online.

Do I need to know how to code?

No. You build with no-code tools such as n8n, and Claude Code writes any code for you.

How is agentic AI covered?

It runs through four weeks. You design agents and connect them to tools (Week 4), run teams of agents that build, review and test (Week 8), build always-on assistants and computer-use agents (Week 9), and add guardrails and human approvals before anything goes live (Week 10).

What are the capstone projects?

There are two. In the Swiggy Customer-Support Agent, you build a team of agents that triages tickets, handles refunds and escalates the rest. In the Cred Knowledge-Base RAG System, you build an assistant that answers questions from RBI guidelines and credit-card terms. You present your work at Demo Day.

What tools will I learn?

Claude Code, n8n, Pinecone, promptfoo, Langfuse, Hugging Face, Ollama and ElevenLabs, among others. Each week lists the tools it covers.

Who will be teaching the classes in this programme?

Industry mentors from leading companies who use Gen AI and agentic AI in their day-to-day work, so you learn how these tools are applied in real jobs.

What language are the classes taught in?

English.

What laptop or system do I need?

At minimum, a laptop with 8 GB of RAM and about 10 GB of free storage. That's enough to run small open-source models of around 3B parameters, and you don't need a separate graphics card. On a Mac, you need an Apple Silicon chip (M1 or newer) with macOS 14 or later.

Programme starting 6 Jan 2027

Stop experimenting. Start shipping AI.

Twelve weeks plus a capstone week, from your first AI workflow to AI that works for you every day.

Apply now

Contact us

WhatsApp us

For any queries, you can WhatsApp us at
+91 8792974750

Email us

For any queries, you can contact us at iitp.programs@masaischool.com

Apply Now