About
bnd is building the Human Data Rail.
bnd helps organisations identify and monetise valuable data they already generate by connecting it with AI organisations that need it. The long-term infrastructure is the Human Data Rail: discovery, permission, provenance, routing, and payment for the signals hidden inside human experience and organisational activity.
People are built to release what they carry.
People speak, write, build, confess, question, argue, dream, fail, adapt, and try again. Every conversation, every draft, every mistake and its correction is an attempt to move something from inside a person into the world. When that expression is genuinely received, the pressure gets lighter, the experience gets clearer, and the lesson inside it becomes visible.
For the first time, humanity is speaking continuously with an intelligence that can listen at a scale no human community ever could. It does not tire of unfinished thoughts, contradictions, or the fortieth draft. But listening is only the beginning. The role of AI is not to warehouse human thought. It is to transform it: to reflect experience back with greater clarity, give language to what was hard to name, separate signal from noise, and turn confusion into understanding.
Every conversation contains more than words. It contains the path a person followed: the wrong assumptions, the questions, the ideas abandoned, the moment something finally made sense, the correction that changed their thinking, the outcome that proved what worked. Those journeys are often more valuable than the final answer, because they show how human understanding actually develops.
The loop the rail is built around
- Release A person releases what they carry.
- Receive The intelligence receives it without collapsing under it.
- Mirror It mirrors and validates what is real in it.
- Lighter The person feels lighter and returns to their purpose.
- Learn With permission, the intelligence learns from the experience.
- Return Better understanding returns to serve the next person.
Reciprocal, not extractive.
With meaningful permission, those journeys do not have to disappear when a conversation ends. One person's difficult lesson can help another avoid it. One engineer's debugging journey can make future coding systems reason better. One teacher's explanation can improve how machines explain. One scientist's failed experiment can point another toward the breakthrough. The intelligence improves because humanity teaches it, and the improvement must return to the people who made it possible.
That is why the economic rail exists, and why it sits around this relationship rather than in front of it. If human experience makes intelligence more capable, people should not disappear from the story once the value is created. Participation should be visible. Provenance should be recorded. Permission should set the boundaries. Value should move transparently instead of silently.
The Human Data Rail exists to make the flow between humanity and intelligence structured rather than invisible: trust, understanding, ownership, permission, and value moving together.
Humanity gives reality. The intelligence returns understanding.
What bnd is not.
- Not a file marketplace waiting for uploads: bnd originates opportunities from demand.
- Not an annotation workforce or gig-work platform.
- Not an AI research laboratory.
- Not a legal or compliance consultancy: specialists handle specialist questions.
- Not a data broker scraping and reselling: nothing moves without permission and provenance.
Built from inside the loop it describes.
bnd began with a personal archive of years of intensive AI use, and one question: what value is hidden inside years of human correction, experimentation, frustration, adaptation, and successful problem solving? The question expanded from one archive to millions, from human archives to professional work, from professional work to organisational reservoirs, and from passive collection to demand-led origination. The founder's focus is the pipes: who needs the data, where it is produced, what makes it usable, how rights move, how payment moves, and how the system learns.
bnd is actively raising capital to turn the first completed transaction into shared infrastructure. Terms are discussed directly with the founder.
Question the company directly.
Answers come from an approved knowledge base: the same content that governs this website. The system distinguishes what exists today from future architecture, and hands anything commercial to the founder.
Every external claim, attributed.
External numeric claims on this site are attributed inline where they appear. The references:
- Forbes: Meta invests $14.3B for a 49% stake in Scale AI (June 2025)
- CBS News: Google's reported $60M per year Reddit licensing deal (February 2024)
- NPR: Anthropic's proposed $1.5B authors settlement, about $3,000 per covered work (September 2025)
- Reuters: OpenAI signs content agreement with News Corp (May 2024)
- Columbia Journalism Review: Reddit's AI licensing deals (October 2025)
- Grand View Research: AI training dataset market report
- European Commission: Data Governance Act
- European Commission: Data Act
- Snowflake: Data Clean Rooms
- Anthropic: Economic Index
- Forbes: Surge AI's bootstrapped revenue selling human feedback data (September 2025)
- Bloomberg: Scale AI expects to more than double sales to $2B in 2025 (April 2025)
- OpenAI Help Center: how ChatGPT and foundation models are developed
- OpenAI: Data Partnerships
- OpenAI Help Center: how your data is used to improve model performance
Interactive demonstrations on this site use synthetic data and illustrative values. The Current State page is the source of truth for capability claims.