Generative AI in Pune

1. Introduction to Python

•  Python Installation and setting up your Environment
•  Get familiar with various Integrated Development Environments (IDEs) such as IDLE, PyCharm
•  Start Python programming by learning to code in an interactive shell.
•  Start programming on an interactive shell.
•   Python Identifiers, Keywords
•  Access Command line arguments within programs.
  
  

2. Conditional Statement, Loops and File Handling

• Python Data Types and Variable  Condition and Loops in Python  Python Modules & Packages
•  Python Files and Directories manipulations
•  Use various files and directory functions for OS operations
  
  

3. Python Core Objects and Functions Built in modules (Library Functions)   Numeric and Math’s Module

•  String/List/Dictionaries/Tuple
•  Complex Data structures in Python  Python built in function
•  Python user defined functions
•  The anonymous Functions – Lambda Functions
•  Decorators
•  Iterators and Iterable  Generators
•  Yield keyword
•  Iterators Generators Difference
  
  

4. Modules & Packages

•  Modules
•  How to import a module…?  
• Packages
•   How to create packages
  
  

5. Data Manipulation with Pandas

•  Data Frames  Series
•  Creating Pandas DataFrame  Selection in DFs
•  Data Describe  Data info
•   Retrieving in DFs
•  Reshaping the DFs – Pivot  Combining DFs
•  Merge
•  Concatenation

6. Introduction to NumPy

•  Array Operations  Arrays Functions
•  Array Mathematics
• Mean
• Standard Deviation  Max
• Min
• Array Manipulation
• Reshaping
• Resizing
• Random function
• Transpose

7. Visualization with Matplotlib

•  Matplotlib Installation
•  Matplotlib Basic Plots & it’ s Containers
•  Matplotlib components and properties
•  Scatter plots  Histograms
•  Bar Graphs  Pie Charts  Box Plots

1. Introduction to NLP, Text Preprocessing Methods

•  Normalization
• Tokenization
• Stopword Removal
• Stemming & Lemmatization
  
  

2. Linguistic and Semantic Analysis

• Part-of-Speech (POS) Tagging
• Named Entity Recognition (NER)
• Sentiment Analysis
  
  

3. Text Vectorization Techniques

•  Bag of Words (BoW)
•  TF-IDF
  
  

4. Hands-on NLP Implementation

• Industry-standard library usage:
• spaCy for pipeline construction
• Hugging Face Datasets for model training/evaluation

1.Transformer Architecture Deep Dive

•   Self-attention mechanisms 
• Encoder-decoder structure
•  Foundation models: BERT (bidirectional), GPT (autoregressive), and LLM variants
  
  

2. LLM Training Lifecycle

•  Pretraining objectives
•  Tokenization strategies: Byte-Pair Encoding (BPE), WordPiece and Embedding
•   Fine-tuning overview
  
  

3. Prompt Engineering Mastery

•  Zero-shot, few-shot, and chain-of- thought prompting
•  Role-based prompting and persona injection
•   Prompt optimization: iterative refinement, constraint specification, and output formatting
  
  

4. Hands-on Model Deployment

•  Hugging Face Transformers
•  Loading pretrained models and tokenizers
•  Text generation with controllable parameters (temperature, top- k/p)
•  OpenAI API integration:
•  Request construction, response parsing, and cost management
•  Safety filters and output validation

5. Application Development

•  Build task-specific agents: summarizers, classifiers, and content generators
•  Evaluate outputs for coherence, factual accuracy, and bias mitigation

1.RAG Architecture Fundamentals

•  Workflow: retrieval → context injection → generation
•  Advantages over pure LLM inference for domain-specific tasks
  
  

2. Embedding & Similarity Search

•  Dense text embeddings (e.g., Sentence-BERT, E5)
•  Vector similarity metrics: cosine, Euclidean, dot product
•  Approximate nearest neighbor search algorithms
  
  

3. Document Processing Pipeline

•  Data ingestion from diverse sources (PDFs, CSVs, web content)
•  Chunking strategies: fixed-size, semantic, and overlap handling
•  Metadata enrichment for filtered retrieval
  
  

4. Vector Database Implementation

•  FAISS for efficient in-memory indexing
•  ChromaDB for persistent, metadata- aware storage
•  Index optimization and hybrid search (keyword + semantic)
  
  

5. Knowledge Bot Development

• Query parsing and intent recognition Context injection techniques:
• Response generation with citation and source attribution
  
  

6. RAG System Enhancement

•  Query augmentation and expansion  
• Re-ranking retrieved results for relevance
•   Evaluation frameworks

1.Multimodal AI Foundations

•  Multimodal datatypes: text, images, audio, video—and their fusion challenges
  
  

2. Key Multimodal Architectures

•  CLIP: contrastive image-text pretraining
•  BLIP: bootstrapped language-image pretraining for captioning
•  Gemini and other unified multimodal foundation models
  
  

3. Multimodal Embeddings and Alignment

• Image-text embedding alignment  
• Cross-modal similarity search and retrieval
  
  

4. Multimodal Application Development

•  Image captioning and visual question answering (VQA)
•  Multimodal chatbots with vision capabilities
  
  

5. Learn Real-World Use Cases

•  Customer support automation with visual context
•  Content creation: text-to-image generation, video description

1.Multi-Agent System Fundamentals

•  Agent definition: autonomy, goals, and environmental interaction
•  Collaborative vs. competitive agent architectures
  
  

2. Agent Design Principles

• Role specialization and responsibility assignment
• Task decomposition and planning
• Agent Communication protocols

3. Multi-Agent Workflow Orchestration Patterns

•  Sequential workflows  
• Parallel workflows
•  Dynamic routing based on task complexity and agent capability

4. Agent Frameworks & Tooling

•  LangGraph for stateful, cyclic agent workflows
•  Microsoft AutoGen for conversational agent teams
•  Task Automation
  
  

5. Advanced Orchestration – Multi-Agent Control Plane (MCP):

• Centralized coordination and scheduling
• Scalable deployment across compute resources
• Resource allocation and load balancing
  
  

6. End-to-End Automation

• Build autonomous systems for research, coding, and business process automation
•  Monitor agent performance, handle
• failures, and enable self-correction

• Master SQL for data querying and management.
• Design REST APIs using FastAPI or Flask.

• Introduction to Git & Distributed Version Control
• Life Cycle
• Create clone & commit Operations
• Push & Update Operations
• Stash, Move, Rename & Delete Operations.

• Explore model monitoring, data drift, and LLMOps with tools like LangSmith,
  MLflow, and Evidently.
• Apply AI to domains like finance (fraud detection), healthcare (patient
  chatbots), insurance (claim processing), HR (resume parsing), government
  (policy summarization), airlines (customer service), and manufacturing (defect
  detection).

• Showcase a full-stack Generative AI Capstone Project tailored to a real-world problem.
• Receive personalized feedback from mentors.
• Explore career paths like Prompt Engineer, GenAI App Developer, or AI Consultant.
• Join an exclusive GenAI alumni community for ongoing support.

• Tools & Frameworks: Python, SQL, Pandas, NumPy, Hugging Face, spaCy, LangChain, OpenAI, FastAPI, Streamlit, LangGraph, AutoGen, MLflow, Azure, Docker, and more.
• Mini Projects: Build a Resume Analyzer, PDF Q&A Bot, and Customer Support Assistant.
• Real-World Case Studies: Apply AI to finance, healthcare, insurance, HR, airlines, manufacturing, and government.
• What You’ll Receive: Weekly assignments, private GitHub access, mini project walkthroughs, a course completion certificate, career guidance, and exclusive alumni community access.