Creators and content are our background.
Before Masana, our team built, owned, and ran a social media network across Instagram and Facebook that reached more than 100 million followers. That meant working with thousands of creators and influencers: finding the right people, briefing them, paying them, and checking what they delivered.
followers reached across our former media network
of creators and influencers worked with
countries our founder has lived and worked in
companies built and sold
Track record from our team's work before Masana.
- Finding the right creators for a brandRecruiting contributors who match your spec
- Briefing creators on what to shootWriting capture guides and running trial recordings
- Reviewing content at volumeChecking every file against your acceptance criteria
- Paying people across many countriesFair, documented contributor pay
- Handling rights and usage termsSigned releases and consent records on file
Why Masana exists.
Models learn from what they see and hear. Too much of that is scraped, reused, and recorded in the same few places.
- Public datasets are shared by everyone, so they stop being an edge.
- Scraped data comes with unclear rights and no consent trail.
- Most training data comes from a handful of countries, so models stumble everywhere else.
- Robots and multimodal models need data that shows real hands doing real tasks, from the right angle, with the right labels.
- We collect that data fresh, to spec, with people who agreed to it and were paid for it.
How we work.
Four rules we hold every project to.
Every dataset is collected for one client's spec. Not pulled off a shelf, not resold to the next buyer unless you agree.
Every contributor signs a clear release in their own language before anything is recorded.
A small batch first, so you judge real data before committing to volume.
Every file is reviewed against the spec. Files that fail are re-captured, not passed along.
Who's behind it.
A small core team of operators and reviewers, with local field coordinators and a contributor network that grows with every project.
Founded by Iman, an American entrepreneur based in Bali who has lived and worked in more than 50 countries. He has built and sold two companies, and owned and ran a social media network that reached more than 100 million followers. Masana grew out of a simple observation: the world AI needs to learn from is far bigger than the data it's trained on.
Based in Bali. Collecting worldwide.
Our deepest network is across Indonesia and Southeast Asia. We take projects anywhere, through our own contributors or local partners, and tell you up front if we can't do it well.
- Home base in Bali, Indonesia.
- Deep coverage across Indonesia and Southeast Asia, including regional languages with little public data.
- Projects elsewhere scoped one by one.
- See where we work
Let's build your dataset.
Tell us what your model needs to see. We'll reply with a scope and a pilot proposal within two business days.
Request a pilot