Built by people who shipped AI for robots
Our team comes from NASA, Google, Amazon, and Ford—bringing deep learning expertise from autonomous vehicles and mobile robotics to geospatial intelligence.
Our Story
Since 2022, Iris Labs (formerly Crescer AI) has been working to speed up geospatial understanding by using artificial intelligence to better analyze 2D and 3D geospatial data.
We started because we'd been users of GIS data for years and saw thousands of people across the globe manually classifying and cleaning geospatial data. We knew there was a better way.
Our background in building deep learning models for mobile robotics gave us a unique perspective: the same techniques that help robots understand their environment in real-time could transform how we process geospatial data at scale.
We're not just building software - we're building production AI systems that work in the hardest environments, from autonomous submarines to defense operations to continental-scale mapping projects.
Our Mission
To be the leading GeoAI partner for organizations tackling complex geospatial challenges
Our Team
Machine learning and computer vision experts from NASA, Google, Amazon, Deloitte, and Ford
Our Location
Based in San Francisco, CA with operations in
Ann Arbor, MI and clients across the globe
Founding Team
Hemanth Sarabu
Co-Founder
Hemanth brings over a decade of experience building AI systems that operate in unstructured, real-world environments.
- • Former head of AI at YC-backed startup
- • ML experience at Google and Symbio Robotics
- • Published researcher at NASA JPL and Georgia Tech Research Institute
- • Advisor to early-stage AI companies
- • Dual master's degrees: Computational Science & Engineering (Machine Learning) and ME (Physics Modeling)
At NASA JPL, his research focused on autonomous navigation and perception for planetary rovers. At Google, he developed computer vision systems for manufacturing automation. His work spans the full stack from algorithm research to production deployment, with particular expertise in few-shot learning and edge AI systems.
Geoff Horowitz
Co-Founder
Geoff's expertise lies at the intersection of computer vision, sensor fusion, and production-ready machine learning systems.
- • Led image-based manufacturing autonomy solutions at Ford Motor Company
- • Developed L2 and L3 autonomous driving systems
- • Holds 19 patents in Robotics and Data Generation
- • Dual master's degrees: Computer Science (Machine Learning) and Mechanical Engineering (Robotics)
At Ford, he led teams developing perception systems for autonomous vehicles - work that required robust performance across diverse conditions and real-time processing constraints. His deep understanding of how to build AI that works reliably in the real world drives Iris Labs' approach to geospatial intelligence, ensuring our models don't just work in the lab but deliver production-grade results at scale.
Expert Team
Our team of machine learning and computer vision experts specializes in generating intelligent, actionable insights from perceptive sensors like LiDAR, cameras, radar, and sonar.
Want to join us?
We're always looking for talented engineers and researchers who want to solve hard problems in geospatial AI.
View Careers