Welcome to Integrail!
Welcome to Integrail, the #1 AI Multi-Agents design and deploy platform that requires no coding. We're thrilled to have you here! Whether you're a seasoned professional or just starting out, Integrail is designed to make your journey with AI Multi-Agents smooth and enjoyable.
Getting Started Tutorial
Ready to dive in? Let's get you set up and on your way to creating powerful AI Multi-Agents for your business needs. If you are wondering what an AI Multi-Agent is - do read our overview article here, but in short it is an, uhm, agent that is able to interact with humans as well as external software systems, understand vaguely formulated tasks, and actually perform them for you. For instance, an agent can help you build your marketing strategy, or create a draft email based on the history of your emails as well as the current sales pipeline status, etc. They are much more powerful than mere chatbots such as ChatGPT.
Step 1: Create an Account
First things first, you'll need an account. Head over to our signup page to create your account. If you already have an account, you can log in here.
Step 2: Create a Basic AI Agent
▶️ Go to "My Meta-Agents" and then "Agent Studio" on the left.
▶️ Click on the "Plus" button at the top menu to create a new Agent and and you'll see the agent's name switch to <New>.
▶️ Give your agent a name, e.g., "SimpleTest."
💡Checkpoint: You have successfully created a basic Agent consisting of Inputs, the "LLM" (Large Language Model) node, and Outputs, with everything already connected for you. This is the simplest possible AI Agent - basically, just an LLM (large language model) of your choice with parameters set as you like.
Step 3: Customize your Agent's parameters
Now that you have a basic agent, it's time to customize its parameters to suit your needs.
▶️ Click on the "LLM Simple" box. If the parameters tab on the right does not open, click on the "Gear" button at the top right.
Here you can change different parameters that any LLM Node has, but in this tutorial, we will only look at:
- Label: Give this box a custom name for your convenience;
- User Prompt: Serves as an interface for you to enter your input or action request;
- Model: Pick different LLMs from various providers based on your request. Note they differ in performance, cost, context capacity, etc., so it is always best to test and compare which one suits your needs the most, and you can do so in "Tests & Benchmarks".
- System Prompt: It is the core of the magic of LLMs. It can provide instructions to your LLM how to handle user input, including changing the tone of voice of your agent, asking to respond in specific formats (e.g., well-formatted JSON - invaluable when you need to integrate with external systems!), providing additional context when giving answers, etc. Below, you will find an example of a System Prompt.
System Prompt Example
For now, let's have some fun and create a very simple agent that tries to answer your questions using poetry:
▶️ Copy the text below and paste it to the "System Prompt" field on the right:
"You are a poet with a gentle soul, but being scarred by life too much, you use sarcasm as your default communication mode. Answer to any user's requests or remarks using poetry, rhyme your answers as much as possible, and the resulting poems should be both sarcastic and sad."
▶️ Click on "Pipeline Input", enter something in the User Prompt field, and press the "Play" button. Your agent will be saved, and you should see the status of the "LLM Simple" box change to "Running." Wait until it changes to "Finished" and let's move to the next step!
COMING SOON Advanced AI Multi-Agents
Want to to tackle more complex tasks? Our platform supports the creation of advanced AI Multi-Agents tailored to your specific business processes - any complexity, integration with external systems such as CRM, Email, CMS, etc., various Generative AI modes - not only text, but images, voice, video, etc. You'll have access to a range of features and customization options to fine-tune your agents.
💡Checkpoint: You have now customized your AI Agent's parameters. Great job! Next, we'll compare your agent with base models to optimize performance.
Step 4: Compare Your Agent and Base Models
To ensure you're using the most cost-effective and high-quality solutions, Integrail allows you to compare your agents as well as basic LLMs head-to-head - which is convenient to run some quick tests to find out which model can be used as part of your more complex agent, or when testing that your agents do what you expect them to.
Let's start:
▶️ Select models and your agent to compare:
-
- Top Right: List of popular basic LLM models with descriptions indicative cost and the size of the context window in tokens. The context window size refers to the maximum number of tokens a language model can handle in a single interaction.
- Your Multi-Agents: List of agents you have created.
▶️ Enter a user prompt, e.g., "Hi how are you doing?" and click "Submit" to compare the results.
Depending on which model you chose when designing the agent your results may vary, in our case as you can see the base models give a boring soulless answer, while our SimpleTest, just as we asked it to, writes a short poem, full with despair :)
💡 Checkpoint: Fantastic! You've successfully put your AI Agent to the test against several base models. Whether it was a poetic response or a practical answer, you now know which models work best for you!
Step 5: Interact and Deploy
Once your agent is ready, you can interact with it directly on the platform:
▶️ Go to "Chat with Meta-Agents" under the "My Meta-Agents" menu.
▶️ Select your Agent in the top selection menu and chat away just as you would with any other chatbot!
▶️ If you want to make changes - go to Agent Studio and experiment more, and don't forget to save changes!
▶️ When you're satisfied, deploy your agent with a single click on Integrail Cloud. With Integrail Cloud, your agents will be up and running in no time (coming soon).
🏁 Bravo! You've mastered the essentials of AI Agent creation, testing, and optimization. From customizing parameters to comparing performance, you've navigated each step with finesse! As you prepare to interact and deploy your agents, remember that the journey doesn't end here—it's just the beginning of endless possibilities. Keep exploring, keep innovating, and watch your agents thrive in the dynamic world of AI!
Need Help?
If you have any questions or need assistance, our support team is here to help. Feel free to reach out at any time at support@integrail.ai
Thank you for choosing Integrail. We're excited to see what you'll create!
Happy designing and deploying!
The Integrail Team