Marketplace

The data the internet doesn't have.

Expert-led speech, video, code, clinical and enterprise data, listed with the numbers. Play a sample, request the set, or bring your own.

29
datasets listed
2.89M+
hours of audio
4.74M+
hours of video
21.1M+
lines of code
9
modalities
Categories

Every set we can license, with the numbers we'd put on a quote.

Library

8 of 29 datasets

Speech & Audio

8
Speech & Audio · PD-001
Multilingual Two-Speaker Conversations

Two-speaker natural conversations recorded in 13 languages, about an hour of sample audio per language.

AUDIO · 44.1kHz · 2CH · 13 LANG · ~13 H
A sanitised excerpt is shared in your data room after a request.
Preview · excerpt on request
Rights review
Request sample →
Speech & Audio · PD-003
Studio Voice-Actor Conversations (Mono)

Simulated conversations performed by consented studio voice actors, recorded mono.

AUDIO · MONO · EN-US · VOLUME TO SPEC
Preview opens once the clip is moved to Perit-owned storage.
Preview · locked
Speech & Audio · PD-004
Licensed English Podcast Catalog (Multitrack)

Long-form English podcast conversations with a separate WAV track per speaker and a mixed M4A.

AUDIO · MULTITRACK · EN · 1,000 H
Per-speaker 3-minute preview opens once the licence is signed.
Preview · locked
Licence in negotiation
Request sample →
Speech & Audio · PD-005
Indic Multilingual Speech (Redacted)

Conversational speech across seven Indic languages and English, with personal details redacted.

AUDIO · MP3 · 7 LANG · 38 CLIPS
A sanitised excerpt is shared in your data room after a request.
Preview · excerpt on request
Rights review
Request sample →
Speech & Audio · PD-011
Perit Duplex Support Conversations (English)

Open full-duplex English support role-play: healthcare scheduling, billing, travel, insurance and IT.

AUDIO · 48kHz · 2CH DUPLEX · EN-US · 15.5 H
A sanitised excerpt is shared in your data room after a request.
Preview · excerpt on request
Open (CC-BY)
Request sample →
Speech & Audio · PD-030
B2B Tech Sales Cold Calls (Indian English)

Unscripted outbound sales calls in Indian English, recorded over telephony, in three length buckets.

AUDIO · TELEPHONY · MONO · EN-IN · 90 CALLS
90calls
~4.6 hunique audio
8 kHznarrowband
Preview · stats only
Rights pending
Request sample →
Bring your own

You're an expert too. You might have data.

Code with a decade of history, calls, lectures, field photos, drawings. If you own it or can license it, tell us the shape and the volume, and we will come back with what it is worth and what clearing it would take.