Course Overview

 Agentic AI: Building Autonomous Systems for Enterprise & Research 

⏱ Duration: 12 Weeks (Approx.) | 48 Hours (Total) | Mode: Online / Blended (Live sessions + Labs) 

Course Description

To provide foundational knowledge and practical skills in investment analysis and portfolio management, enabling informed decision-making and effective portfolio strategies.

Instructor(s):

• Lead Instructor: Professor Dr. Emdad Khan | Research Professor, Maharishi International University (MIU), USA.
• CEO, InternetSpeech Inc. (25+ Years Experience in AI, Holder of 23 US Patents).

Course Description

This course equips students with the advanced skills to design, build, and deploy autonomous AI agents. Moving beyond simple generative content creation, participants will learn to build systems that can reason, plan, and execute complex workflows autonomously. Students will master Agentic architectures, memory management, tool integration, and multi-agent coordination using industry-leading frameworks.

Evidence of Demand

The AI industry is shifting from "Generative AI" (content creation) to "Agentic AI" (task execution). Tech giants like Microsoft (AutoGen) and OpenAI (Assistants API) are investing heavily in agents. There is a massive skills gap for professionals who can build autonomous systems that reason, plan, and use tools to automate complex workflows in enterprise and academia.

Purpose and Objectives

To equip students with the skills to design, build, and deploy autonomous AI agents.

Objectives:

• Understand Agentic AI architecture & cognitive loops.
• Master advanced prompting (ReAct, Chain-of-Thought).
• Implement Multi-Agent Systems using LangChain/CrewAI.
• Deploy agents for real-world automation (Academic/Admin).

Course Content & Class Plan (Modules)

• Module 1: Foundations of Agentic AI (Evolution, Characteristics).
• Module 2: Agent Architectures (Reactive vs. Deliberative, BDI).
• Module 3: LLMs as Agents (Reasoning Engines, Prompting).
• Module 4: Planning & Reasoning (CoT, ReAct, Self-Correction).
• Module 5: Memory & Knowledge (Vector DBs, RAG, MCP).
• Module 6: Tool Use & Action (Function Calling, API Integration).
• Module 7: Multi-Agent Systems (Coordination, Swarm Intelligence).
• Module 8: Frameworks (LangChain, AutoGPT, CrewAI).
• Module 9: Evaluation, Safety & Ethics (Reliability, Governance).
• Module 10: Deployment & Scaling (Monitoring, Security).
• Module 11 & 11(A): Applied Use Cases (Academic, Admin, Customer Service) & Reliable Grounding.
• Module 12: Capstone Project (Design & Live Demo).

Practical & Field Work (Capstone Projects)

Students will build fully functional agents such as:

• AI Academic Advisor Agent: Reads university policies and advises students.
• Department Service Automation Agent: Handles student tickets and schedules autonomously.
• Multi-Agent Research Assistant: A team of agents (Researcher + Writer + Editor) that writes reports.
• Autonomous Helpdesk Agent: Resolves technical queries by accessing manuals and databases.

Learning Outcomes

Upon completion, students will be able to:

• Understand Agentic AI principles and architectures.
• Design autonomous agents with goals, memory, and tools.
• Implement single and multi-agent systems in Python.
• Integrate agents with real-world APIs and databases.
• Evaluate agent reliability and safety.
• Deploy live agentic systems for enterprise use cases.

Target Audience & Requirements

Target Audience:

• CSE/SWE/CIS/ITM/MCT Students (3rd/4th Year).
• Software Engineers & Data Scientists.
• Automation Engineers & Researchers.
• Faculty interested in AI automation.

Entry Requirements:

• Minimum Age: 20 Years.
• Introduction to AI / Machine Learning.
• Basic Python Programming.
• Fundamentals of Data Structures (Recommended).

Career Pathways

• AI Agent Architect.
• AI Automation Engineer.
• Research Engineer (AI/ML).
• AI Product Developer.

Assessment Criteria

• Final Project – 40%
• Labs & system design tasks – 30%
• Case studies & assignments – 30%

Tools & Resources

• Language: Python.
• Environment: VS Code / Jupyter.
• Models: OpenAI API, Open Source LLMs (Llama/Mistral).
• Databases: Vector DBs (FAISS, Pinecone, ChromaDB).
• Frameworks: LangChain, CrewAI, AutoGen.
• Integration: REST APIs, GitHub.

Financial Information & Certification

• Tentative Course Fee:
• For Daffodil Students Tk 5,000/= 
• Without Daffodil Students Tk 7,500/=.