AI Agent Development with LLM (Build you first agentic AI): Level 1

• Environment Setup
• Data Type
• Variable & Constants
• Data Types
• Type Conversion
• Input/Output Operations
• Operators
• Conditional Statements (if-else)
• Loops (for. while)
• Functions
• Lambda Functions
• List Comprehension
• String Manipulation
• List
• Tuple
• Dictionary
• Set
• Exception Handling
• File Handling
• Module & Package
• Virtual Environment
• pip & Dependency Management

নেবেন: Abdullah Al Noman

• Extraction Techniques from various Sources
• Transformation Techniques from Various Sources
• SQL (How to Deal with SQL Source)
• Feature Engineering
• Loading techniques into various destination
• Task Scheduler in ETL Automation
• Build a ETL Automation Workflow using Airflow
• Build a projects

নেবেন: Salman Md Sultan

• Lowercasing
• Removing Punctuation
• Tokenization
• Stopwords Removal
• Stemming
• Lemmatization
• Removing Numbers
• Removing Extra Whitespace

নেবেন: Salman Md Sultan

• Bag of Words (BoW)
• TF-IDF
• CountVectorizer
• Word Embeddings (After Deep Learning)
• Sentence Embeddings (After Deep Learning)

নেবেন: Salman Md Sultan

• Artificial Neural Network
• Deep Neural Network
• Convolutional Neural Network
• RNN
• LSTM
• GRU
• 2 Projects

নেবেন: Salman Md Sultan

• Transformer
• Attention with Self Attention
• Positional Encoding & Layer Normalization

নেবেন: Salman Md Sultan

• AI Agent vs Normal Chatbot — পার্থক্য
• LLM কী? — tokens - context window - embeddings (concept)
• Prompt Engineering Basics
• System / Instruction / Persona prompt design
• ReAct / Chain-of-Thought — ধারণা
• Hallucination — কেন হয় & কিভাবে কমাই
• Case Study: ChatGPT style agent কীভাবে কাজ করে

নেবেন: Salman Md Sultan

• API basics — REST - keys - rate limits
• OpenAI/Gemini/Claude API overview
• Function Calling/Tool Usage
• Error handling & fallback
• Logging & debugging
• langchain
• pydantic

নেবেন: Salman Md Sultan

• Conversation memory concepts
• Short-term vs Long-term memory
• Vector DB intro
• Embeddings concept
• RAG basics
• Knowledge-base agent build
• Hallucination reduce with RAG

নেবেন: Salman Md Sultan

• Multi-step reasoning
• Task decomposition(planner–executor)
• ReAct+Tool use
• Multi-agent communication
• State management
• Failure recovery
• Research assistant agent

নেবেন: Salman Md Sultan

• What is MCP(Model Context Protocol)
• Why MCP for Agents
• mcp server install & setup
• mcp.json config
• Connect agent with MCP
• File/Web/DB server examples
• Tool discovery & execution
• Debugging & best practices
• Agent demo with MCP

নেবেন: Salman Md Sultan

• Customer support agent
• Document Q&A agent
• Code assistant
• WhatsApp/Telegram/Web chatbot
• Lead generation agent
• Data analysis agent
• Deployment options

নেবেন: Salman Md Sultan

• Safety & guardrails
• Prompt injection defense
• Cost optimization
• Latency & caching
• Hybrid/offline embeddings
• Evaluation metrics

নেবেন: Salman Md Sultan