Question 1

What problem does Agentic AI solve that a regular LLM can't?

Accepted Answer

A plain LLM call is stateless and single-shot you give it a prompt, it gives you text back, and it has no way to check whether that text actually solved your problem.
- It can't look anything up, take an action, or try again if it's wrong.
- Agentic AI solves this by wrapping the model in a loop: the system can call tools, observe results, and keep iterating until the actual goal not just a plausible-sounding answer is achieved.

Example: Ask an LLM "what's the cheapest flight to Delhi tomorrow" and it will guess based on training data or say it can't know. An agentic system will call a live flights API, compare real prices and give you an answer that's actually correct today.

Question 2

How does Agentic AI differ from traditional AI?

Accepted Answer

Traditional AI reacts to input or follows fixed rules. Agentic AI adds planning, reasoning, tool use and memory, so it can take initiative and complete multi-step tasks unsupervised.

Traditional AI
- Traditional AI is designed to perform specific tasks such as prediction, classification, recommendation, or automation.
- It follows predefined rules or learned patterns to generate outputs.
- It typically responds to a single input without independently planning future actions.
- It requires user intervention for each new task or decision.
- It is commonly used in spam detection, recommendation systems, fraud detection, and image classification.
- Goal: Solve a specific problem or automate a well-defined task.

Agentic AI
- Agentic AI is designed to autonomously plan, reason, and execute multi-step tasks.
- It can break complex goals into smaller tasks and decide the sequence of actions needed.
- It interacts with external tools, APIs, databases, and software systems to complete tasks.
- It adapts its actions based on intermediate results and feedback.
- It requires minimal user intervention after the goal is specified.
- Goal: Achieve a user-defined objective by reasoning, planning,

Question 3

What's the difference between AI Agent vs Agentic AI?

Accepted Answer

AI Agent and Agentic AI are closely related but not the same. An AI agent is a building block; Agentic AI is the overall system architecture.

AI Agent
- An AI Agent is a software entity that performs specific tasks on behalf of a user or system.
- It perceives its environment, processes information, and takes actions to achieve a defined objective.
- It can use tools, APIs, databases, or other software to complete tasks.
- It is usually designed for a particular domain or workflow.
- Multiple AI agents can work together to solve complex problems.
- Goal: Complete a specific task or objective autonomously.

Agentic AI
- Agentic AI is a broader AI paradigm that emphasizes autonomous reasoning, planning, and decision-making.
- It enables AI systems to break complex goals into smaller tasks, adapt to changing conditions, and execute multi-step workflows.
- It often consists of one or more AI agents working together with memory, planning, and tool usage.
- It can continuously evaluate progress and modify its actions based on feedback.
- It is used to build highly autonomous systems capable of achieving complex goals.
- Goal: Build AI systems that can independently reason, plan, and acc

Question 4

What's the difference between Agentic AI vs Generative AI?

Accepted Answer

Generative AI creates content from a prompt. Agentic AI achieves goals by reasoning, planning and acting, often using a generative model as its "brain."

Generative AI
- Generative AI is designed to create new content such as text, images, audio, videos, and code.
- It generates responses based on patterns learned from large training datasets.
- It typically responds to user prompts without independently planning multiple actions.
- It is widely used for content generation, summarization, translation, image generation, and code generation.
- Examples include ChatGPT for text generation and image generation models like DALL·E.
- Goal: Generate high-quality and contextually relevant content.

Agentic AI
- Agentic AI is designed to autonomously plan, reason, and perform multi-step tasks to achieve a goal.
- It can make decisions, use tools, interact with external systems, and adapt its actions based on feedback.
- It breaks complex problems into smaller tasks and executes them sequentially.
- It often integrates LLMs with tools, APIs, databases, memory, and planning mechanisms.
- It is widely used for AI assistants, workflow automation, software development agents, and research assist

Question 5

What are the different types of AI agents?

Accepted Answer

Classified by how much reasoning and memory they use:
- Simple reflex agents: Act on current input only, using fixed rules, no memory.
- Model-based reflex agents: Keep an internal world model to handle partial information.
- Goal-based agents: Choose actions that move toward a specific goal.
- Utility-based agents: Pick the action that maximizes a "utility" score among valid options.
- Learning agents: Improve over time from feedback.

Question 6

What are the primary components of AI agents?

Accepted Answer

An AI agent generally consists of five main components:
1. Memory: It helps the agent recall past actions, user preferences and context.
2. Reasoning Engine: This part decides what the agent should do next based on logic or goals.
3. Tool Use: Agents use APIs, databases or other services to perform specific actions.
4. Communication Layer: Enables interaction with users and other systems.
5. Environment Interface: Allows agents to perceive and modify their surroundings.

Question 7

Why is memory critical for the performance of AI agents?

Accepted Answer

Memory gives an agent context, continuity and the ability to learn — without it, every query is treated as a brand-new conversation.
- Short-term memory: Context within one conversation/task.
- Long-term memory: Persistent knowledge, often in a vector database.
- Working memory: Supports reasoning during a multi-step task.

Example: An assistant with memory recalls you prefer meetings after 2 PM and applies that automatically next time.

Question 8

What's the difference between Agentic System vs User Prompt?

Accepted Answer

An Agentic System is an AI system that autonomously plans, reasons, and executes tasks to achieve a goal, whereas a User Prompt is the instruction or query provided by the user that tells the AI what to do.

Agentic System
- An Agentic System is an AI system capable of reasoning, planning, and executing multi-step tasks autonomously.
- It can break a complex goal into smaller tasks and determine the sequence of actions.
- It can use external tools, APIs, databases, memory, and search engines to complete tasks.
- It adapts its actions based on intermediate results and feedback.
- It requires minimal user intervention after receiving the objective.
- Goal: Achieve a user-defined objective through autonomous planning and execution.

User Prompt
- A User Prompt is the input, instruction, or question given by the user to an AI model.
- It specifies what the user wants the AI to do or answer.
- It serves as the starting point for the AI's response or task execution.
- The quality and clarity of the prompt influence the quality of the AI's output.
- It does not perform reasoning or actions by itself; it only provides guidance to the AI.
- Goal: Communicate the user's intent or request to

Question 9

What are tools in the context of Agentic AI?

Accepted Answer

In Agentic AI, tools refer to external functions, APIs or software utilities that an AI agent can call to perform actions beyond simple text generation. They allow the agent to interact with the real world, execute commands, retrieve live data and complete complex workflows.
- Searching the web for current information.
- Querying a database to fetch or update records.
- Running code for analysis.
- Sending an email or calendar invite.

This mechanism where the model invokes a function with structured arguments is called function/tool calling.

Question 10

Common frameworks for building Agentic AI systems?

Accepted Answer

Frameworks provide ready-made building blocks for planning, memory and tool orchestration instead of writing an agent loop from scratch.
- LangChain: Chains reasoning steps, integrates tools.
- LlamaIndex: Connects models to structured/unstructured data.
- CrewAI / AutoGen: Multi-agent collaboration.
- ReAct: A prompting pattern that interleaves reasoning with action, not a full framework.