Next-generation agentic systems that reason, adapt, and endure in the real world.
We build continuously operating, multimodal agentic AI right at the intersection of agentic and physical AI. Our neuro-symbolic stack combines deep neural perception, empirical reasoning, adaptive control, and long-term learning, enabling autonomous systems to perceive, decide, act, and improve under real-world uncertainty.
Alongside our robotics work, we contribute to open-source agentic systems through OmegaClaw, a neural-symbolic agent framework built on the Hyperon AGI stack, MeTTa, and open-source technology from the ASI Alliance ecosystem.
OmegaClaw unifies large language models with a formal symbolic layer to create stateful cognitive architectures capable of long-term persistence, auditable reasoning, autonomous goal management, and tool use.
We created the initial OmegaClaw prototype and a MeTTa-based core with self-programming capability. This is a key founder-led open-source contribution that is aligned with the company roadmap.
Oma is the first Telegram agent built on the OmegaClaw framework. Interacting with Oma is the fastest way to experience what we are building with OmegaClaw.
NARTECH is our reasoning-first AI technology: a neuro-symbolic reasoning engine operating on top of state-of-the-art deep learning for perception.
Our breakthrough: an effective neuro-inspired inference control mechanism focuses compute on what matters, in real time.
NARTECH is our reasoning-first AI technology: a neuro-symbolic reasoning engine operating on top of state-of-the-art deep learning for perception.
For robotics, NARTECH integrates seamlessly with ROS 2 perception and control stacks, enabling adaptive autonomy in real environments.
Experience our technology in real-world environments: unknown terrain, uncertainty, and on‑the‑fly improvisation. Featuring robust transfer to unfamiliar environments, with decisions you can inspect and trust.
Reasoning capabilities showcased in complex domains with easy integration into other applications featuring high real-time perception throughput.
Our agentic systems work builds on Hyperon, MeTTa, and open-source technology from the ASI Alliance ecosystem.
Testing autonomous reasoning systems in real urban environments with the Aarhus City Lab team.
View post →
Interview on NARS-based reasoning-learning systems and real-time autonomy.
Read full story →
Partners driving advanced reasoning tech into healthcare, manufacturing, and urban infrastructure.
Read more →
An AI assistant, Trusted and Explainable Artificial Intelligence for Saving Lives (TruePAL), to provide real-time warning of risks of potential crashes to the first responders.
Read report →
NASA’s TruePAL platform brings real-time situational awareness to improve road-side safety for officers.
Read article →
Temple University’s longstanding NARS (Non-Axiomatic Reasoning System) project—a real-time, experience-learning AGI system—continues its pioneering work, now partnering with Future AI.
Read full story →Real-time empirical reasoning on a mobile robot; integration with perception and control for robust autonomy.
Neural perception + symbolic reasoning for collision risk estimation with interpretable warnings.
Framework unifying experience-based reasoning with learning for sample-efficient, adaptive behavior.
We collaborate with academia and industry.
AI researcher focused on neuro-symbolic learning, real-time reasoning, robotic autonomy, and persistent agentic systems.
Cognitive scientist focused on adaptive control, multimodal perception, and embodied autonomy.