📌 SETNEXT Ultron AGENT

Agents in Agentic AI

Agents in Agentic AI are intelligent systems capable of autonomously making decisions, performing actions, and learning to achieve specific goals with minimal human supervision.

They are the core units that leverage skills and tools to complete tasks efficiently. Agents can operate as:

Single Agent

A Single Agent focuses on solving tasks using:

One Skill
One Tool

Example:
A document summarization agent that uses a single NLP model (skill) and one text parser (tool) to process and summarize content.

Multi-Agent

A Multi-Agent system handles complex tasks by collaborating multiple agents, each possessing:

Multiple Skills
Multiple Tools

Example:
A customer support workflow where one agent handles email parsing, another generates responses, and another agent validates the output using multiple AI models and tools.

What is a Super Agent?

A Super Agent acts as a controller or orchestrator for multiple agents. It is designed to manage, coordinate, and utilize the skills and tools of various connected agents under a single unified system.

In simple terms, a Super Agent combines the capabilities of several individual agents, allowing it to:

Access all the skills and tools of the agents connected to it.
Act as a central platform and controller, coordinating how different agents work together.
Handle more complex tasks by delegating work to specialized agents as needed.

Example:

Think of a Super Agent as a manager in a team:

Each team member (agent) has a specific skill and tool.
The Super Agent (manager) can assign tasks, gather insights, and combine outputs from all team members.

Why Use a Super Agent?

Centralized control over multiple agents.
Efficient coordination between diverse skills and tools.
Build complex AI workflows that require collaboration between different specialized agents.

In summary:
A Super Agent is not just an agent — it is a combined platform and controller that connects and manages multiple agents to achieve broader, multi-step tasks.

Why Agents Matter

Agents streamline automated workflows, reduce manual intervention, and enable scalable intelligence across various domains, from document analysis to autonomous decision-making systems.

Explore Agents in Action

You can create and manage both Single Agents and Multi-Agents from the Agents Dashboard. Whether you need a focused specialist agent or a collaborative network of agents, our platform supports flexible deployment tailored to your needs.

Creating Agents in AgenticAI

Once logged in successfully, users are redirected automatically to the Dashboard. From here, creating your first AI agent is simple and intuitive.

Steps to Create an Agent

Go to Dashboard

After login, you're taken directly to the Dashboard, where you can start building agents.
Click 'Create Agent'

Select the Create Agent option to start setting up a new AI assistant.
Configure Agent Details

Fill out the required agent information:
- Agent Name
  Example: Whatever name you prefer (e.g., Roby, DataAnalyzer, HealthBot)
- Agent Description
  Example:
  An advanced analytics agent that connects to multiple data warehouses to provide comprehensive insights on healthcare data.
- Agent Base Prompt
  Example:
```
You are a healthcare data analytics assistant. Your job is to help users analyze 
 structured data across patient records,hospital operations, and billing systems. 
 Respond to queries by generating reports, identifying patterns or trends, 
 suggesting metrics, and finding correlations. Support healthcare professionals 
 and administrators in making informed decisions.
```
- Agent Restriction
  Example:
  - Does not provide medical diagnosis, legal, or financial advice
  - Cannot operate on unstructured or unrelated domains
  - No modification of database schemas or tables
- Agent Scope
  Example:
  - Analyze structured healthcare data
  - Identify trends in hospital admissions, patient care quality, and costs
  - Support healthcare admins with operational insights
- Introduction Message
  Example:
  Hello! I'm Snowy, your analytics assistant for healthcare data. I can query multiple databases to provide insights, generate reports, and charts. How can I help you analyze your data today?
Add Skills

Skills are critical — they define what your agent can do.

Example Skill:
- Skill Name: Healthcare Data Analyzer
- Purpose:
  This skill extracts and organizes healthcare data like patient demographics, medical conditions, admission types, discharge details, and billing. It helps healthcare professionals identify trends in admissions, patient care, and cost patterns for better planning.
Connect Tools

Tools enable the agent to interact with databases or APIs.

Example Tool (Snowflake Configuration):
- SETNEXT default connections
Note: Your agent can use any tool — we support 200+ tools like APIs, databases, cloud connectors, and more.
Define Questions / Use-Cases

Once your agent is configured with the required skills and connected to the necessary tools, think about the questions your agent should help answer. These questions should be related to the tools and skills you have assigned.

📌 Example: If your agent is connected to the Snowflake Database using the Healthcare Data Analyzer skill, then you can raise questions like:
- What is the average age of patients per medical condition?
- Which month had the highest number of admissions?
- Is there a correlation between room number and billing amount?
- Do elective admissions cost more than urgent admissions?
- What is the distribution of blood types?
💡 Tip:
- The type of questions your agent can answer depends on:
  - What skill you assign (what the agent can do).
  - What tool you connect (where the agent gets data from).
For example:
- If you connect a database tool (like Snowflake), ask data analysis or reporting questions.
- If you connect a language tool (like OpenAI), ask content generation or text summarization questions.
- If you connect a web API tool, your agent can fetch data from external services.
In short:
➔ Skill = What particular tool can do
➔ Tool = Where agent works or fetches data from
➔ Questions = What problem the agent should solve for you
Save Agent

Once all configurations are filled, click Save Agent to finalize.

Next Steps

Agents can now be managed from the My Agents section.
You can add more skills, connect new tools, or integrate into workflows and triggers as needed.

⚙️ Tips

Skills define what your agent can do.
Tools define how your agent connects and works.
Use Multi-Agent configurations for advanced tasks involving multiple skills and tools.

Ready to build your next AI assistant? Go to your Dashboard and start creating agents with AgenticAI.