KIVA AI Rebrands to Perle, Bringing Human Wisdom to AI Models
KIVA AI today announced that it has rebranded as Perle, to deliver modular training data solutions that bring human wisdom to AI models. Perle takes the heavy lifting out of training and fine-tuning complex multi-modal AI models, giving customers powerful tools for everything from defining project requirements, RLHF, and expert human data tagging, to data enrichment, cleaning, and defending against adversarial attacks. By delivering actionable insights and intelligence guided by an expert-in-the-loop, Perle allows customers to focus on accelerating innovation. Learn more about Perle or request a demo at www.perle.ai.
Perle’s first training solution is available now and delivers expert human data annotation via a vetted global network of domain experts spanning industries as diverse as STEM, Legal, and Linguistics. The company will also release an innovative AI project scoping feature in the coming weeks that directly addresses challenges with defining annotation project scope and guidelines in order to create custom optimized workflows. Additional offerings will roll out in the coming months.
Many AI teams spend the majority of their time and resources on undifferentiated data management rather than model development. Companies can spend up to 80 percent of their development resources on data labeling, discovery, quality checks, compliance, versioning, bias auditing, drift detection, and more. Teams often are also forced to choose between expensive one-size-fits-all platforms they don’t need, or wrestle with a patchwork of services from different sources. Perle solves these challenges with a modular training data solution that helps companies choose the right mix of services to bring higher-quality AI models to market faster, and transform their organizations to focus on innovation over administration.
“Perle is designed to help companies of all shapes and sizes extract ‘Perles of Wisdom’ from their data and accelerate innovation with their AI model development with our expert-in-the-loop solutions,” said Ahmed Rashad, Founder and CEO of Perle. “We’re obsessed with making it easy for customers to innovate by taking the heavy lifting of data management off their plates, and empowering them to choose the solutions they need to build better models faster and more efficiently.”
Founded by a team of AI veterans from companies and institutions including Scale, Amazon, Meta, MIT, Harvard, UC Berkeley, and Columbia, the company raised $8.5 million in pre-seed funding in October 2024.
Last Updated on 2 days by News Desk 1
