AI Agent Guide

To get started with the Neuralis AI Ecosystem, head over to our website and hit 'Get Started' or 'Go to App'. From here, there's a guide on how to start creating AI Agents in the Training Center section. This guide, which will be extended to include details on the upcoming Public Marketplace and AI Agent Teams section, is also detailed below. Should you have any questions about how to use any of our functionalities or if you are missing any options or configurations, please reach out to [e-mail].

---

Accessing the Platform Accessing the platform can be done by clicking on the 'Create Agent' button on the left hand side, or by navigating to this link. From here, you have a set of configurations to choose from, of which all are currently free of charge. There will be a number of paid optionalities in the future, allowing users to create more complex and popular AI Agents.

---

Choosing a Framework In this first step, users can choose the underlying AI framework that will power their agent. Each framework offers different capabilities and integration options. As of May 2025, the current options are:

1. ElizaOS Our native framework. Provides the best integration with the Neuralis platform and offers balanced capabilities for most use cases.

2. CrewAI Specialized in multi-agent collaboration. Ideal for complex workflows requiring coordination between multiple AI entities.

3. LangChain Robust framework with extensive tool integrations. Best for applications requiring sophisticated chains of reasoning and external tool usage.

4. OpenAgents Community-driven framework with a focus on openness and customization. Great for specialized agents with unique requirements.

---

Selecting a Model After choosing a framework, users are asked to choose one of our models. The current options, as of May 2025, are stated below and include the most popular GPT models.

• GPT-3.5 Turbo

• GPT-4o Mini

• GPT-4o

• GPT-4 Turbo

The upcoming, paid models, include the below:

• Claude 3

• Gemini Pro & Ultra

• Mistral AI

• OpenSource Models (i.e. Llama)

---

Character Details Want your AI Agent to act and talk in a certain way, and preset with certain professions or knowledge? In this section, you can configure a number of important characteristics, including:

• Agent Name & Description: Give your agent a distinct identity with a name that reflects its purpose. Write a detailed description of what your agent does and how it should behave.

• Professions: Select up to 3 professional domains your agent should specialize in. This helps optimize its knowledge and responses for specific subject areas.

• Personality Traits: Adjust the sliders to configure your agent's communication style

• Adjectives: Select up to 6 adjectives that best describe your agent's personality. These help shape the tone and approach of your agent's interactions.

• Agent Image: Upload a custom image to represent your agent in the interface. This helps visually identify different agents in your dashboard.

---

Add-Ons: Coming Soon With the upcoming 'Add-Ons', users can enhance their agent with specialized capabilities. These add-ons extend the agent's functionality beyond basic conversation. This section is an upcoming paid functionality and can include some of the below preliminary add-ons.

• Code Interpreter Enables your agent to write, analyze, and execute code in various programming languages.

• Data Analysis Allows your agent to analyze and visualize complex datasets and draw insights.

• File Access Enables your agent to read and write files for document processing and data management.

• Web Search Gives your agent the ability to search the web for up-to-date information.

• Network Access Allows your agent to access network resources and external APIs for enhanced functionality.

• Blockchain Integration Connects your agent with blockchain networks for data retrieval and interaction.

• On-Chain Price Feeds Real-time price data from blockchain oracles for financial and trading applications.