microagi deploys robots in industrial operations through Atlas, capturing customers' manual work, training task-specific models, and integrating suitable hardware. Its Shift marketplace gathers first-person human demonstration data for physical AI.
Notable milestones from microagi’s journey — funding rounds, launches, and other moments worth knowing.
MicroAGI partnered with Google Cloud to run its robotics training and inference workloads on NVIDIA Blackwell infrastructure. The collaboration pairs its task-specific Atlas platform with multimodal data processing and rack-scale compute for commercial and industrial robot deployments.
SourceHummingbird led the round, joined by Northzone, LocalGlobe, Village Global and redalpine. The funding is intended to scale Atlas across large industrial companies, roughly ten months after microagi was founded.
Atlas scopes factory tasks, captures how workers perform them, trains a suitable model, and integrates hardware without locking customers to one vendor. Its data engine had passed 100,000 hours of real-world work at launch.
Shift launched in New York with free professional apartment cleaning in exchange for first-person recordings from camera-wearing cleaners. The footage captures how people perform physical tasks and becomes training data for robotics and AI labs.
Eight months after starting in Munich, microagi opened a research headquarters at Bahnhofstrasse 63 and made Zürich the center of its global research programs. The company chose the city over San Francisco for its concentration of robotics talent and proximity to European industry.
microagi began with five people in a Munich hacker house, founded by engineers from Red Bull Racing and Mercedes-AMG Petronas. The team brought Formula 1's focus on reliable machines under pressure to the problem of deploying physical AI.
Source
