In March 2024 Perle was founded with a simple, but audacious goal - bring human wisdom to artificial intelligence. We believe that AI is only as smart as the data that shapes it. That’s why we’re focused on creating modular solutions that put human experts in the loop to help companies effortlessly train their multi-modal AI models faster, more affordably, and with higher-quality data, annotated by a global network of human experts.

As we reflect on our first year, it’s humbling to look back on how much our team has accomplished together, and equally exciting to consider what’s on the horizon.
 

March 2024: The Birth of Perle

Originally founded as KIVA AI, Perle was the brainchild of founder and CEO Ahmed Rashad.  After years building AI solutions at companies including Scale and Amazon, Rashad realized that training data solutions were too bloated, too slow, too expensive, and too light on human expertise. Low-quality annotations, inconsistent workforce engagement, and opaque compensation models plagued AI development. 

Perle was born to change that. The team got to work building a modular annotation solution that prioritizes expertise, engagement, and fair compensation—laying the foundation for a radically different way to generate high-quality AI training data. 

‍
October 2024: Raising $8.5M Pre-Seed Round

The venture capital community quickly took notice, and by October Perle had closed $8.5M in pre-seed funding to help the company grow and scale. We’re grateful for our bench of VC and Angel Investors who believe in our vision and provide invaluable support and insight along the way. We’re thrilled to be on this journey together.

‍

January 2025: Launching Our MVP

In just seven months we moved from concept to execution, launching an MVP to prove our core vision: human experts produce higher quality data for training AI models. By putting the solution into the hands of customers, we were encouraged by the early results–higher-quality data, faster iteration speeds, and a highly engaged global network of human experts.

By prioritizing quality, operational efficiencies and fairness, we delivered an alternative to the status quo of leveraging low-skill gig workers to train a company’s most important models.

‍

February: Rebranding from KIVA AI to Perle

We kicked off 2025 with a bang, officially launching our data annotation following its successful MVP period.  To celebrate the launch, we also introduced the world to our new branding, changing the name of the company from KIVA AI to Perle.

The name Perle is taken from the expression “Pearls of Wisdom.” This reflects our belief that hidden deep within endless piles of data are the highly polished gems that are truly useful, and can only be surfaced and polished by human experts.

‍

March 2025: First Team Offsite

Like many companies, Perle was founded as a global, virtual entity.  We built the team from the best talent in the world, and embraced a virtual working environment.  A few weeks ago the entire team got together for the first time in Austin, TX to meet IRL, take stock of the last year, and chart the course for the year ahead. 

It was an incredible opportunity to celebrate with each other and acknowledge what we’ve accomplished in such a short time. Like AI, companies really thrive when they’re rooted in these powerful human connections and insights.

‍

The Road Ahead

Perle is just getting started.  While we’re in the market with our first solution focusing on data annotation with experts in the loop, we’re hard at work on a suite of modular solutions that help AI researchers and developers at every stage of developing and training their multi-modal AI models.
‍

In the coming months, look for news about an AI assistant to help scope complex training projects, tools to help clean and enrich data, and solutions to help address adversarial threats.
‍

At Perle, we’re committed to bringing human wisdom to AI making it effortless for companies to tap into the highest quality data, faster and more affordably than ever before. It’s been a thrilling ride so far and we can’t wait to show you what we’re working on next!

‍