Hirundo

We’ve found a way to store it efficiently and inexpensively creating Green hydrogen

visual

About Us

Hirundo is pioneering next-generation AI accuracy for mobility. Our platform automatically detects and remediates mislabeled and low-quality data across radar, LiDAR, vision, and other sensor modalities – dramatically improving the performance of AV, ADAS, and other mission-critical models.

Unlike conventional tools such as manual QA, Hirundo uses a patented Data Influence Engine to analyze how each training sample affects a model’s behavior. The result is precise, automated data QA and optimization that works in minutes – not weeks. Our platform is domain-agnostic, fully airgap-deployable, and scales seamlessly across datasets and model types.

We’re backed by five filed U.S. patents and led by a team of former elite Israeli R&D officers and academic leaders. Today, Hirundo is in active collaborations with the Israeli Security Agency, ELTA, and global mobility enterprises.

From optimizing data quality pipelines to ensuring real-world robustness, Hirundo makes your AI more accurate, trustworthy, and production-ready.

Relevant Challenges

Automated Systems

Hirundo addresses the core challenge of automated systems: model fragility due to bad data. In autonomous vehicles and ADAS, mislabeled or low-quality data leads to unreliable perception and decision-making. Hirundo’s platform automatically identifies and fixes these issues, enabling mobility teams catch and correct errors in radar, LiDAR, and vision datasets without manual review.

Data Privacy & Protection

Hirundo‘s platform also enhances data protection and security in mobility by enabling precise, on-premise removal of sensitive or non-compliant data from both datasets and trained models. Our platform can detect and eliminate personally identifiable information (PII), confidential sensor data, and other high-risk content – without retraining. This is critical in mobility, where data flows from vehicles, passengers, and infrastructure must meet strict privacy and regulatory standards

Smart Manufacturing

Hirundo improves AI accuracy in smart manufacturing by automatically detecting and correcting mislabeled or faulty data in visual inspection and sensor-based QA systems. This leads to more reliable defect detection, fewer false positives, and faster deployment of factory AI systems.

Year Founded:
2022
Country:
Israel
Funding Stage:
Seed
Product Stage:
Pilot to First Paying Customer
Sector:
Mobility Services, Vehicle Technologies
Geographic Focus:
North America (USA/ Canada), Europe
Introduce me

Contact EcoMotion to learn more about this startup and their activities

Contact Us
Skip to content