Building distributed AI systems shouldn't mean choosing between cutting-edge capabilities and battle-tested reliability. Today, we're excited to announce the release of Genesis—an open-source Python framework that brings the robustness of industrial middleware to the world of AI agents.
AI agents are transforming how we build intelligent systems. But deploying them in production—especially across distributed networks—introduces familiar problems: service discovery, reliable communication, fault tolerance, and scalability. Traditional AI frameworks often rely on fragile REST APIs or message brokers that become single points of failure.
Meanwhile, industries like aerospace, healthcare, and autonomous vehicles have solved these distributed systems challenges for decades using the Data Distribution Service (DDS) standard. What if we could bring that proven reliability to AI agents?
Genesis bridges these two worlds. Built on RTI Connext DDS—the same middleware powering surgical robots, flight control systems, and autonomous vehicles—Genesis provides a framework where AI agents discover each other automatically, communicate with sub-millisecond latency, and operate without central brokers or complex configuration.
Start your agents and services anywhere on the network. They find each other automatically—no IP addresses, no port configuration, no service registries. Add new capabilities by simply starting new components.
Agents can call other agents as naturally as they call functions. Your PersonalAssistant agent discovers a WeatherAgent, and the LLM sees it as just another tool to invoke. Chain agents together dynamically based on what's available at runtime.
Don't overwhelm your LLM with hundreds of available functions. Genesis classifiers automatically select the 5-10 most relevant functions for each query, reducing token usage by 90%+ while improving accuracy.
Switch between OpenAI, Anthropic, or add your own LLM provider with minimal code changes. The framework abstracts provider differences so your agent logic stays clean.
Genesis is designed for teams building AI systems that need to work reliably:
- Robotics and autonomous systems requiring real-time AI decision-making
- Enterprise deployments where uptime and scalability matter
- Research teams exploring multi-agent architectures
- Anyone tired of fragile microservice configurations
Genesis is available now under the RTI License. Whether you're exploring with a 60-day evaluation or building with Connext Express, you can have your first multi-agent system running in minutes.
git clone https://github.com/rticommunity/rti-genesis
cd rti-genesis
./setup.sh
cd examples/MultiAgent
./run_interactive_demo.shVisit our GitHub repository to dive in, or reach out to genesis@rti.com with questions.
The future of AI is distributed. Genesis makes it reliable.
Genesis is developed in collaboration with RTI, the infrastructure software company for physical AI systems.
(c) 2025 Copyright, Real-Time Innovations, Inc. (RTI) All rights reserved.
RTI grants Licensee a license to use, modify, compile, and create derivative works of the Software. Licensee has the right to distribute object form only for use with RTI products. The Software is provided "as is", with no warranty of any type, including any warranty for fitness for any purpose. RTI is under no obligation to maintain or support the Software. RTI shall not be liable for any incidental or consequential damages arising out of the use or inability to use the software.