Introduction to AgentRunner
AgentRunner is a cutting-edge platform designed to streamline the development and deployment of AI agents. Born from the challenges of building AI solutions, AgentRunner addresses the pain points experienced by software companies, developers, AI engineers, prompt engineers, and AI enthusiasts alike. Our platform allows you to visually engineer AI workflows, iterate rapidly, and manage versions with ease, all while ensuring scalability and integration with your existing software. AgentRunner solves the problems that arise during software development involving AI flows. It's designed to be quick and easy, handling AI workflows, iterations, logs, versioning, API creation, and prompt flow building.
AgentRunner was internally developed to solve the founders issues during building AI solutions in their software. The initial software, which aimed to automate content variety and ad creation, faced significant hurdles. The traditional development process, which involved writing prompts, testing them, and then handing them off to developers, proved slow and ineffective. The resulting AI integrations often failed to meet expectations. The team recognized that building AI agents and workflows demanded a more flexible and iterative approach than traditional code development. This realization led to the creation of AgentRunner, a platform designed to handle AI workflows separately from the core code, allowing developers to focus on API calls while AgentRunner manages the AI logic.
Key Features and Benefits
Visual AI Workflow Engineering
AgentRunner's node editor provides a visual interface for designing and iterating on AI workflows. This approach eliminates the need for hard-coding prompts and logic into the software, making it easier to create iterations of your flow much more smoothly than through programming.
Drag-and-Drop Interface: Build complex AI workflows with an intuitive, user-friendly node editor.
Real-Time Iteration: Quickly test and refine your AI agents without waiting for lengthy development cycles.
No-Code Approach: Focus on the logic of your AI workflows without needing extensive coding knowledge.
Comprehensive Logging and Monitoring
AgentRunner meticulously logs every step of your AI agent's execution, allowing you to identify and resolve issues quickly.
Detailed Run History: Track inputs and outputs for every API call and AI agent run.
Node-Level Insights: Examine the inputs and outputs of each node in your workflow to pinpoint areas for improvement.
Efficient Debugging: Quickly identify and correct issues within complex AI workflows.
Version Control and Collaboration
AgentRunner's versioning system allows you to manage and iterate on your AI workflows without disrupting your live applications.
Branching and Merging: Create new versions of your AI workflows and merge them seamlessly.
Stable Versioning: Designate a "stable" version for production use while continuing to develop and test new iterations.
Collaborative Development: Enable multiple team members to work on different versions of the same AI workflow simultaneously.
Scalability and Integration
AgentRunner is built on GoLang, a high performance programming language, designed to scale your AI solutions.
API Integration: Connect AgentRunner to any software through APIs, enabling seamless integration with your existing systems.
App Store: Access a variety of pre-built apps and integrations to extend the functionality of AgentRunner.
Custom Apps: Add your own private or public apps to the AgentRunner app store for personalized workflows.
Getting Started with AgentRunner
Prerequisites
Before you begin using AgentRunner, ensure you have the following:
A valid AgentRunner account. Visit agentrunner.com to sign up.
An API key, if you plan to use AI models.
Basic knowledge of AI concepts and workflow design.
Step-by-Step Guide
Access the AgentRunner Platform:
Go to agentrunner.com and log in to your account.
Create a Project:
Click on the "Create New Project" button.
Enter a name and description for your project.
Create a Workflow
Click on the "Add a Flow" button and select "Create New".
Name your workflow.
Design Your Workflow:
Use the node editor to drag and drop the required components.
Connect the nodes to create the desired flow of data.
Configure each node with the appropriate settings, such as API keys and model parameters.
Integrate with your Software:
Make an API call from your software, passing the necessary variables.
Receive a response back from AgentRunner with the output from your AI agent.
An example CURL API call is available within your workflow editor by clicking on the API button.
Monitor and Debug:
Check the logs for every step of the API call or AI agent run.
Examine the inputs and outputs of each node to identify any issues.
Create Versions:
Version your AI workflow to create a new version of your software.
Select a version and create a new one, splitting it into a new version.
Edit the new version without making changes to the original.
Expected Outcome
By following these steps, you should be able to successfully create, deploy, and manage AI agents using AgentRunner. You will be able to iterate on your workflows, monitor their performance, and integrate them seamlessly into your existing software.
Beta Program and Future Plans
AgentRunner is currently in beta, and we encourage software companies to try it out and provide feedback. During the beta period, AgentRunner is free to use, allowing you to explore its features and benefits without any cost. We are actively integrating with multiple software platforms to make AgentRunner even more versatile and accessible. As we continue to develop AgentRunner, we are committed to complying with the highest data security policies and best practices in online development. Our goal is to acquire SOC 2 certification during the summer 2025 to ensure the highest standards of data protection.
While in beta, please use AgentRunner for test projects only, as unexpected changes and new versions may occur frequently. Our policies are still being finalized during this phase.
Understanding AI Agents
An AI agent is essentially a workflow that involves AI solutions to complete a specific job. Think of it as capturing the step-by-step instructions of an expert and automating them to achieve the highest quality outcome. You provide the necessary inputs, and the AI agent processes them to deliver the best possible result in a matter of seconds or minutes.
Support and Resources
For additional support and resources, please visit:
Documentation: help.agentrunner.com
Videos: Available within the AgentRunner app.
AI Chatbot: Ask our AI chatbot for assistance anytime.
Contact Information
AgentRunner is owned and operated by GBD Software as a Service Limited, based in Hungary.
Website: agentrunner.com
We are excited to see what you build with AgentRunner!
