Liberty Source PBC provides human-in-the-loop data services that deliver high-accuracy labeling, annotation, and testing for AI and machine learning applications, particularly in autonomous systems and language model fine-tuning. By employing a US-based workforce, the company ensures data security and compliance while enhancing model performance through precise data preparation and quality assurance.
Funding
$910K raised to dateRaised to date based on public sources. This may differ from the amount the company actually raised and is based only on what is publicly available on the internet.
Founders
Product
Problem
AI and machine learning models, particularly in autonomous systems and language model fine-tuning, require high-quality labeled, annotated, and tested data. Achieving high accuracy in these data services is challenging, especially when dealing with complex edge cases and the need for secure, compliant data handling.
Solution
Liberty Source PBC, now part of Digital Divide Data (DDD), provides human-in-the-loop data services, delivering precise labeling, annotation, and testing for AI and machine learning applications. The company specializes in multimodal data for autonomous systems, including LiDAR, multi-sensor data, and object tracking, as well as language model fine-tuning, RLHF, and red teaming. By utilizing a 100% US-based workforce, Liberty Source ensures data security and compliance while achieving over 99.5% accuracy across data pipelines. Their AI-integrated tools and subject matter experts enable the delivery of reliable training data solutions, adapting to clients' existing technology stacks.
Target Audience
The primary customers are organizations in the defense, federal, and public sectors, as well as Fortune 500 companies and autonomous driving companies, requiring high-accuracy data for AI and machine learning applications.
Features
- Multimodal labeling and annotation for autonomous vehicles, robots, and drones, including LiDAR and semantic segmentation.
- Expertise in GEOINT AI, annotating 2D/3D, video, satellite & drone imagery, and GPS data.
- RLHF and fine-tuning services for language models, including RAG enhancement, model evaluation, and prompt engineering.
- Red teaming services to identify vulnerabilities, biases, and harmful outputs of generative models.
- Data engineering and data analysis services, including data integration, visualization, verification, and migration.
- AI-assisted quality checks and operational excellence to deliver 99.5%+ accuracy.
- SOC2 Type 2 certification and adherence to NIST 800-705 standards for data security.
- Platform agnostic approach, integrating with clients' existing technology stacks.