Indri Labs vs Troveo: licensed data network or sourcing service?

What is Troveo?

Troveo is a Texas-based company that licenses real-world data for AI training. It launched in November 2024 with $4.5M in seed funding led by Seven Seven Six, the firm founded by Reddit co-founder Alexis Ohanian, with angels including Zynga founder Mark Pincus. Its founder and CEO is Marty Pesis, who previously founded Vouch.

It started with creator and studio video. In April 2026 it added audio, text, enterprise workflow data, gaming and robotics, and said it had paid more than $20 million to content owners. On its website Troveo says it holds 8M+ hours of video and 4M+ hours of audio, 1,000+ gaming titles, 350+ company datasets and a network of 7,000+ data providers across 150+ countries.

We are Indri Labs, so read this page knowing that. Everything said here about Troveo comes from its own website and press coverage, listed under sources, as of October 2026.

How does Troveo work for AI labs?

Troveo offers buyers three routes:

• A dataset library you can filter by video, audio, gaming, robotics, image and text, with listings such as sports commentary and multilingual conversations
• Company datasets: exports from business systems such as CRM, support tickets, email, chat and finance tools, with names, credentials and client identities removed under what Troveo calls a documented anonymization standard
• Sourcing: you share a brief and Troveo says it will source, produce and annotate data to your spec, combining programmatic classification with human validation

Troveo does not publish prices. Its sourcing page says the data team responds to requests within 24 hours.

How does Troveo pay creators and rights holders?

For video owners, Troveo says it licenses content to AI labs on their behalf under NDA with no public listings, that owners keep full ownership, and that every use earns a royalty. It prefers 1080p or 4K footage the owner fully holds rights to, and says processing takes one to two weeks. Semafor reported in October 2025 that, according to Pesis, a handful of creators had been paid more than $1 million each.

For companies, Troveo says it handles export, anonymization and packaging, that ownership never transfers, and that it pays on every sale, not just the first. It does not publish a revenue share percentage. Indri Labs also helps companies sell their data to labs; see licensing your data with Indri Labs.

How do Troveo and Indri Labs compare?

Both license data from its owners and both will collect to a spec. The difference is where they start: Troveo from a large network and library, Indri Labs from your description of the need.

How does Indri Labs work?

Indri Labs is a service, not a marketplace. You describe the need: modality, volume, coverage across regions, languages and edge cases, format, rights and deadline. We scope it for free and send a one-page brief that names and prices every source, includes a sample from each, gives the total data cost and shows our fee on its own line. You can walk away owing nothing.

If you approve, you sign one contract and pay one invoice, with us. We negotiate a license with every seller, usually companies that have never sold their data, and record what doesn't exist yet, such as first-person video on smart glasses and head cameras filmed to spec. You receive one merged dataset with duplicates and personal data removed, checked against your spec, with the source and license for every record, in your format.

We source first-person video, operational logs (telemetry, transactions, system records), calls and meetings as audio with transcripts, and documents. Sensor and robotics data and clinical records are coming soon. More on egocentric video data.

When is Troveo the better choice?

• You need large volumes of licensed video or audio, such as broadcast, sports or conversational speech, and an existing library covers it
• You want gaming data with synchronised keystrokes, which Troveo lists and Indri Labs does not source
• You need text corpora from publishers, which is outside Indri Labs' current data types
• Speed matters more than a custom spec, and data that already exists will do

When is Indri Labs the better choice?

• The data sits inside companies and is not listed anywhere
• You need first-person video recorded on smart glasses or head cameras to your spec
• You want every source named and priced, with a sample from each, before you commit
• You want several sources merged into one dataset under one contract and one invoice

For a wider view of the market, see AI training data providers compared. If the data you need is not in a library, send us the need. The brief is free.