Six products. The full record.

The end goal · direction, not shipped

Law within reach

The end point is to make the law easily accessible: your rights in the palm of your hand, so you can gain the knowledge, fight your own case, prepare documents professionally, and save time.

The same approach could one day help speed up places where a backlog of cases has to be checked one by one, such as parts of government and the courts and tribunals system. A slow system prolongs the damage. Many people end up accepting injustice because the process is built to wear ordinary people down, while the companies and people who caused the harm carry on untouched. We want to shorten that distance.

The end goal · direction, not shipped

One place to read the sky, every way the world did

The end point is for everyone to try the different cultural readings of the sky in one place, and to learn how each one is calculated and derived: how people across the world saw the sky and the stars, and connected them to personality. Learning about yourself, and your connection to the stars.

Later we want to bring in ancient books, palm leaves and old papers, in one place, added by the people who work with them: researchers, scholars, archaeologists and anyone connected to the work, each with its source recorded.

The end goal · direction, not shipped

Bring the night-time story back

We used to sit at night to share horror stories, and sit with children for a bedtime story before sleep. We want that back. It also preserves the beliefs and generational stories of old times: people who fought for their land and their freedom and became the legends of a place.

Stories carry values. As the world moves on with information shaped by social media and the news, many people lose track of what is right and what is wrong, and what is worth fighting for. They are easily baited by people who lack morals and values, and put the wrong people in power. Governments can be at fault, but the people who put them there are citizens. Stories that teach courage, fairness and whom to trust are part of how a society chooses well.

Stories also set the imagination running. Many of our own ideas began with a story: a simple cartoon that sent the mind wandering, or a conversation where someone shared what they knew and we noted down our ideas. The Mahabharata, the Ramayana and the Bhagavad Gita do this at scale. They show a way of life, and through the stories of kingdoms, ministers, teachers, gurus, sculptors, mathematicians and others, they record how people wrote down their views and knowledge and shared them with the world. They open a door: if you walk this path, the end can look like this. They also show what is worth learning from these people, and what to avoid.

The story map ahead

Only The 12 Jyotirlingas is live today. These are stories we are hunting for. Each one will be cited to its source, and its rights will be researched and recorded before it ships: old tales may be free to tell, but a particular retelling, translation or photograph may not be.

The end goal · direction, not shipped

Your own chef, at home

More people than ever live alone, and many choose to order food in rather than cook. Dish is for the bachelor, the beginner and anyone who does not know what to make: to cook healthy food from what is already in the kitchen.

The feeling we are aiming for is that you have your own chef in your own kitchen.

The end goal · direction, not shipped

Can models find leaner circuits?

The end point is to take advanced models and test whether they can reduce the number of resistors, transistors, capacitors and other components in a circuit while keeping it working, which raises efficiency, and whether they can propose genuinely new designs and structural changes in electrical circuits.

If that works, it could feed back into the boards and modules people build with, the Arduino-class microcontroller boards and the sensor and power modules around them. Every result will be checked by a real circuit solver, never taken on the model’s word.

The end goal · direction, not shipped

Design yourself

For everyday users: know what suits you for a meeting, a date or a function, using the clothes you already own instead of being sold new ones.

For the end goal: cleanly custom-design your own outfit in our software. Cosplayers who hesitate because they think they must hire a designer and pay for trials could try a character themselves. Someone interested in fashion design who could not enrol on a course gets a practical, live simulation to learn on.

Beyond that: designing yourself rather than a generic 3D model. If a VR game that scans you and turns you into a usable character arrives, this becomes the way to build a realistic one, and producing a game is a stated final goal of this project.

Measured by us

What our own experiments proved, and why the hurdle is high

What we proved

• The signal is real. At 391 people, the link between what the camera reads and the human skin-tone rating is about 6.4 standard deviations from zero (r = -0.33). It is weak, about 11% of the variation, but it is not noise.
• We know where the variation comes from. Within one fixed take it is 2.52 L*; between different people 9.31; between captures of the same person 11.25. The same face varies more with the lighting than people vary from each other, which is a measured reason to control the capture instead of trying to model the light.
• Combining captures helps. Using every capture of a person instead of one lifted the link by about a third (0.221 to 0.296).
• Undertone holds steadier than depth. In one face over five years, depth swung across almost the whole scale while undertone held to a spread of about 5 degrees. One person, so a lead and not a claim, and still untested across devices.
• Every shortcut around the lighting failed at proper size. Colour-cast correction (classic and learned), exact shading removal, an eye-white reference and more frames were each tested and did not rescue a casual photo. That is why controlled capture is the route, and not just a preference.

Why the hurdle is high

• A casual selfie reads differently each time: a second session gives a different three-way answer about 40% of the time.
• The end game needs a match within about 1 to 2 delta-E. With the camera, light, session and exposure all fixed, our readings still vary by about 2.5 L*. Across sessions it is 11.25, and across 22 seconds of casual phone photos we saw a 37-point swing. We are roughly an order of magnitude away. Published work reaches about 1.5 delta-E only with a physical colour target in the frame.
• So the road to the end game is controlled capture with a colour reference in frame: a booth first, then a kit people hold up, and not a casual app. Cosplayers are the audience most willing to do that.