Today marks the day Kled enters new territory: 54 New Multimodal Datasets.

Since we started 4 months ago we’ve had the pleasure to work directly with top AI labs in Silicon Valley. We’ve had a front row seat on exactly where AI is heading and what are the most important consumer datasets to collect.

So we’re pleased to announce that on top of our image and video collection efforts we’ve identified and built out interfaces to collect 54 new multimodal datasets from the Kled user base.

Some examples include:

• Voice Clarity Dataset: Training AI to understand speech in noisy environments. Crucial for sales call agents, customer service bots, and real-time transcription.

• OCR + Receipt Dataset: Millions of real receipts and documents to help AI read handwriting, parse numbers, and unlock automation for accounting, fintech, and expense tracking.

• AR/VR Interaction Dataset: Users navigating AR/VR with gestures + voice. Building the foundation for spatial computing, gaming, and immersive AI copilots.

• Household Task Dataset: Short video clips of everyday chores and tool use. Essential for robotics, home assistants, and embodied AI models.

We’ve amassed over 20,000 trained VAs who will begin to contribute to these data sets until we’ve fully refined the collection interfaces and can push them to the main Kled app, fully accessible at anyone’s fingertips. Overall this should greatly increase our pipeline of buyers and strengthen all our enterprise and AI lab relationships. As well as provide new monetary opportunities to all of our users.