A multi-agent system is a computer system which is composed of multiple intelligent agents. Multi agent system can solve problems that are difficult or impossible to solve by individual agents or a monolithic system. Developers face many difficulties when they try to develop multi- agent systems, for example, passing messages over networks, manage agents' life cycles based on agent type.
There are few other platforms which help to solve this problem such as SPADE, JADE, NetLogo. Even though these frameworks are useful in academia and small applications they are difficult to use in production environments due to scalability and maintainability. Therefore we suggest a platform which enables us to develop multi-agent systems with distributed agent development. We called it Rakun.
Rakun is a distributed Multi-Agent Platform. Using rakun developers can build their own Multi-Agent systems. Here rakun works as an agent controlling software with ability to manage each agent's life cycles such as Start, Stop, Restart, etc. This helps developers to manage their Multi Agent system without worrying about deployments or scalability. Also rakun agents come with the ability to work as distributed agents, this also helps Multi-Agent System to run without interrupting even though there are few agents not working. Our main goal is to provide a powerful solution to Multi-Agent System developers to build their systems.
- Distributed Agents
- Async message passing and handling
- Fault Tolerance
- Cross platform
- Live monitoring
- Websocket enabled
- Drag and drop agent development
- Web dashboard
- Allow developers to deploy agent system based on pay as go method