Luel

Founded by Inigo Lenderking in 2025

Access multi-modal datasets to train machine perception models that listen, reason, and act

Pricing model

Paid

Time to value

Quick Setup (< 1 hour)

You can use Luel to access specialized multimodal datasets and build foundational machine perception systems. The laboratory focuses on training models to process audio and visual inputs, reason about complex environments, and execute actions. Through its continuous data engine, Luel collects and curates diverse training data from worldwide contributors. Researchers and machine learning developers can browse available datasets, contribute new sensory data, and utilize foundational models to build real-world applications that require computers to perceive, understand, and interact with physical surroundings.

What Luel does

Browse multi-modal perception datasetsContribute sensory data to the continuous engineTrain models on combined audio and visual inputsEvaluate foundational machine perception modelsDownload structured datasets for AI researchCollect distributed multi-modal data from global contributorsContinuous proprietary data engine for ongoing dataset updatesMulti-modal training data combining audio and visual inputsPublic dataset browser for machine perception researchersCommunity contribution framework to collect new sensory inputsFoundational models focused on listening, reasoning, and actingMachine perception benchmarks for autonomous systems