Luel
Enterprise-grade data supply

Curated datasets for production AI teams

License ready-to-ship datasets or brief us on a custom build. Every collection is rights-cleared, quality audited, and delivered with enterprise support.

Browse catalog

What’s inside

We source from vetted contributors, maintain consent logs, and cross-check every file for duplicates, safety issues, and instruction compliance.

  • Structured delivery

    JSON manifests with clip metadata, transcripts, QA scores, and direct S3 download links.

  • Compliance on autopilot

    Consent releases, PII audits, and audit logging baked in for every dataset.

  • Custom augmentation

    Need annotations, translations, or balance adjustments? We deliver end-to-end.

Browse curated datasets

Filter by use case or modality to find the best fit for your training needs.

Use Case
Modality
VideoAudioSensor DataCustom
Egocentric Vision for Accessibility AI

1,400 hours of first-person video from blind and low-vision users with rich accessibility metadata

Clips

5,000

Hours

1400

Use Cases

Accessibility Tech, Computer Vision

  • Real-world POV from blind/low-vision users across diverse daily scenarios
  • Rich metadata: OCR outputs with bounding boxes, conversation transcripts, and head IMU motion data
  • Hands and objects clearly visible for manipulation understanding and object recognition
Explore dataset
AudioTextCustom
Nighttime Traffic Audio Narrations

Urban walk-and-talk narrations with layered ambient noise profiling

Clips

28,800

Hours

1200

Languages

en-US

Use Cases

Transportation, Urban Planning

  • Noise labels include spectral centroid & SPL readings for ML filtering
  • Multiple mic placements (handheld, lapel, smartphone pocket)
  • Narration topics cover transit delays, ride share pickups, public safety
Explore dataset
VideoAudioTextCustom
POV Kitchen Cleanup Sessions

First-person kitchen cleanups with narration, action labels, and bounding boxes

Clips

180

Use Cases

Smart Home, Robotics

  • Multi-angle coverage: primary POV plus optional static wide shot
  • Action narration with on-screen captions for recognition and alignment
  • Tool and ingredient metadata (brand, material, observed state)
Explore dataset
AudioTextCustom
Multilingual Contact Center QA

Dual-channel bilingual agent and customer calls with full QA metadata

Clips

2,400

Hours

160

Languages

es-MX, es-US, en-US

Use Cases

Customer Support, Telecom

  • Accent-balanced agent/customer pairs for code-switching and bilingual intent modeling
  • Diarized dual-channel WAV @48kHz with separate metadata for agent vs customer
  • Scenario pack covers refunds, loyalty points, device activation, and outage escalations
Explore dataset
AudioTextCustom
Healthcare Intake & Triage Voices

Near-field and far-field nurse/patient triage dialogues with empathy scores

Clips

1,200

Hours

45

Languages

en-US

Use Cases

Healthcare, Telehealth

  • HIPAA-safe synthetic cases vetted by clinical advisors
  • Turn-level intent tags (chief complaint, follow-up, escalation)
  • Acoustic diversity: quiet home office, waiting room ambience, vehicle cabin recordings
Explore dataset
Ready when you are

Need something bespoke? Let’s scope it together.

Whether you’re training speech-to-text models, vision systems, or multimodal assistants, Luel's collection network and QA pipeline plug directly into your roadmap.

  • End-to-end vendor — sourcing, QA, legal, delivery
  • Flexible licensing models — flat fee, per minute, or revenue share
  • Dedicated account team for enterprises with recurring needs

Let’s work together

Tell us a bit about the data you need and how we can help.