Video summary

GEN AI & AGENTIC AI with Python - Session -01| Ashok IT.

Main summary

Key takeaways

Educational

Main ideas / lessons conveyed

  • Purpose of the course: Help learners transition from being AI users (using tools like ChatGPT/Copilot) to AI developers (building AI applications similar in capability to those tools), using Python.
  • Why this matters now: The speaker claims companies increasingly want developers who can build GenAI/agentic AI systems, not just people who can “prompt” or use AI tools.
  • Core skill focus: Python is presented as the key language for building LLM-based applications, RAG systems, and agentic workflows.
  • From basics to real projects: The course is positioned as “zero to hero”, requiring no prior prerequisites (including for non-IT and non-programming backgrounds).
  • Market-aligned stack: The training is mapped to popular tooling/concepts such as LLMs, prompt engineering, RAG, vector databases, LangChain, LangGraph, MCP, ML Ops/LLM Ops, and deployment/production readiness.
  • Career outcomes: After completion, learners should be able to apply for roles like:
    • Data Scientist
    • AI/ML Engineer
    • GenAI / Agentic AI Developer
    • RAG Developer
    • AI Application Developer
    • AI Tester
    • MLOps engineer
    • Python developer / LLM application developer

Methodology / learning approach (as described)

  • Start with foundational Python
  • Learn required Python libraries for AI/ML/GenAI
  • Cover ML/DL fundamentals before GenAI
  • Introduce LLM concepts and prompt engineering to use tokens effectively and improve output quality
  • Build GenAI applications using RAG
  • Build agentic AI systems using frameworks/tools:
    • LangChain for pipelines
    • LangGraph for more complex graphs/decision-making loops
    • MCP for integrating multiple systems (servers/clients) in agent workflows
  • Move into production deployment concepts via ML Ops / LLM Ops
  • Include interview preparation + resume building
  • Emphasize “real-time projects” and practical coding sessions throughout

Detailed course content / modules

Prerequisites

  • No specific prerequisites required
    • Suitable for beginners
    • Works for:
      • Non-IT backgrounds
      • Freshers
      • Java/.NET/DevOps backgrounds
      • People without programming experience

What you will learn (high level)

  • Python from scratch
  • Python libraries needed for AI development (examples mentioned):
    • NumPy
    • Pandas
    • Matplotlib
    • FastAPI (exposing as REST APIs)
    • PyTorch
    • TensorFlow
    • Streamlit (UI)
    • Scikit-learn
    • Probability & statistics
  • Machine learning fundamentals
    • Supervised / unsupervised / reinforcement learning
    • Regression, classification, clustering
    • Decision trees, random forest
  • Deep learning fundamentals
    • Backpropagation
    • Optimizers
    • CNN, RNN, Transformers
  • LLMs and GenAI
  • Prompt engineering
  • RAG systems with vector databases + embeddings
  • AI agents / agent workflows (agentic AI)
  • Real-time project implementation
  • Cloud deployment + ML Ops / LLM Ops
  • Interview preparation + resume building
  • Additional mentioned topics include:
    • Git/GitHub and version control
    • Dockerization
    • Kubernetes
    • CI/CD pipelines
    • Logging and production-ready configuration
    • Deployment automation (mentions “N8N architecture”)

Module breakdown

  1. Module 1: Python programming (Core + Advanced)

    • Python intro, installation, environment setup
    • Variables, operators, conditionals, loops
    • Strings, data structures
    • Functions, packages/modules
    • File handling, exception handling
    • OOPs
    • Working with APIs
    • Database connectivity
    • Real-time coding practices

  2. Module 2: Python libraries for AI development

    • NumPy (numerical Python)
    • Pandas (data cleaning)
    • Matplotlib (visualization)
    • Probability & statistics
    • Scikit-learn (ML projects)
    • FastAPI (REST API exposure)
    • Streamlit UI
    • (Earlier references also include PyTorch / TensorFlow)

  3. Module 3: Machine learning + deep learning algorithms

    • ML: supervised/unsupervised/RL, regression/classification/clustering
    • Algorithms: decision trees, random forest
    • Deep learning: backpropagation, optimizers
    • Models: CNN, RNN, Transformers

  4. Module 4: LLMs + prompt engineering

    • Explains LLM basics and how tools like ChatGPT/Copilot work (front-end vs LLM in background)
    • Key concepts mentioned:
      • Prompt vs token concepts
      • Context window, temperature, parameters
      • LLM integration and development
    • Prompt engineering techniques (explicitly named):
      • Zero-shot prompting
      • One-shot prompting
      • Few-shot prompting
      • Chain-of-thought prompting
      • Step-by-step prompting
      • Role-based prompting
    • Prompt templates, prompt best practices
    • Integrations mentioned:
      • OpenAI model integration
      • Google Gemini integration

  5. Module 5: Generative AI application development with RAG

    • Text generation application development
    • Chatbot development
    • Document Q&A (incl. PDF-based QA)
    • Embeddings + vector databases
    • RAG architecture and implementation
    • Build RAG apps using:
      • PDFs, websites, databases
    • Knowledge-base driven chatbot using LangChain pipelines

  6. Module 6: Agentic AI development (LangChain, LangGraph, MCP, agents)

    • Build intelligent AI agents capable of:
      • Planning
      • Thinking
      • Using tools
      • Completing tasks autonomously (multi-step execution)
    • Agent planning, agent workflows, tools usage, multi-step tasks
    • LangChain for pipeline-style agent development
    • LangGraph for complex flows requiring loops/conditionals
    • MCP servers/clients integration for agent environments
    • Build “own agents” using the combined approaches

  7. Module 7: ML Ops and LLM Ops

    • Deploy and maintain AI apps as production-grade systems
    • Focus on cloud deployment/operations for ML/LLM applications

Course logistics and deliverables (as stated)

  • Start date: “starting from today” (first session)
  • Class timing: 7:00 p.m. to 8:15 p.m. IST
  • Days per week: Monday to Friday (5 days/week)
  • Duration: 3 months
  • Mode: Online live classes
  • Fee: ₹25,000 for 3 months
  • What students receive:
    • Daily live classes
    • Soft-copy materials
    • Class recordings
    • Recording access validity: 1 year after course completion
    • Real-time projects development
    • Interview preparation + resume building
    • Practical coding sessions
    • AI tools exposure
  • Additional note: Google Classroom used to share notes/videos immediately after class

Job roles mentioned (post-training)

  • Data Scientist
  • AI Engineer
  • ML Engineer
  • GenAI Developer
  • Agentic AI (Agent-DKI) Developer (spoken as “Agent-DKI”)
  • AI Application Developer
  • AI Tester
  • Python Developer
  • LLM Application Developer
  • RAG Developer
  • AI Automation Developer
  • ML Ops Engineer (explicitly mentioned)

Q&A highlights included in the subtitles

  • Probability & statistics: Yes, covered (as part of libraries)
  • NLP coverage: Yes, will be covered
  • Real-time projects: Yes, for every concept (estimated ~10 projects)
  • EMI options: Yes (offered)
  • Cloud integration (Azure/AWS): Yes (cloud deployment covered; integration implied)
  • Anaconda in Python: Yes, will be covered
  • Assignments: Daily tasks will be provided
  • Resume “experience years” guidance (for experienced learner):
    • Suggestion: can highlight roughly last ~2.5 to 3 years worth on resume (while keeping the individual’s prior experience context)

Speakers / sources featured

  • Mr. Ashok (trainer/founder of Ashok IT) — primary speaker and course instructor.
  • Participants mentioned by name in the Q&A portion:
    • Yogesh
    • Gafur
    • Meghana
    • Madhusudan
    • Ashish (called out as “Ashish sir”)
    • Ravi (asked multiple questions)
    • Bhushan
    • Parvinder
    • Tejas
    • Saurav
    • Jeddu
    • Vinayak
    • Madhu Kumar
    • (Additional names may appear but were not fully clear in the subtitles.)

Original video