About Tashvi
Modern creativity for an ancient craft.
Tashvi is built for the jewelers, designers, and custom-order studios who actually do the work, by people who grew up in the trade.
Why we built it
Tashvi started in a family jewelry store. Watching custom orders come in (engagement rings, anniversary pieces, redesigned heirlooms) was watching the heart of the craft. It was also watching where the work broke down.
Custom design moves slowly because the tools haven't moved. A client describes a ring. A CAD designer charges by the hour to interpret it. Weeks of revisions follow. The client struggles to picture the piece until the wax sample is in their hand. Designers spend their best hours translating, not designing.
Made better. Made faster. Made for one.
By 2024, generative AI had become good enough for almost everything else (writing, code, photography, video), but jewelry had been left behind because solving it required two worlds that rarely overlap: machine learning and the bench. So we built the tool we wished the family store had ten years earlier. Tashvi is what happens when AI finally meets a craft on its own terms.

The founder
Charles Nischal
Charles worked at Microsoft as a software engineer in Azure cloud and AI, watching generative models grow from research curiosities into working tools. The years there taught him what it takes to ship machine learning into the kind of systems people rely on every day.
He is also a Diamond Graduate of the International Gemological Institute, one of the world's largest gemological labs. The program covers the 4Cs in depth, grading under loupe and microscope, fluorescence behavior, treatment detection, and the documentation standards used across the trade.
After Microsoft, he joined his family's jewelry retail store. He saw what custom orders cost in time and revisions, watched designers spend their best hours interpreting instead of designing, and noticed how often clients walked out because they couldn't picture what they were buying.
Tashvi brings those three threads together: engineering rigor, formal gemological training, and a working knowledge of the trade. It is a design tool built by someone who has stood on both sides of the counter.

What we believe
Manufacturing-ready, not just pretty pictures.
Anyone can train an AI to make beautiful jewelry images. Few train it to respect prong depth, band thickness, stone proportion, and setting geometry that actually casts. Tashvi outputs are renders a bench jeweler can work from, not concepts they have to reinterpret.
Built with jewelers, not for an imagined jeweler.
Every word in the AI's vocabulary (pavé, bezel, cathedral, milgrain, toi-et-moi) comes from working studios. The roadmap is shaped by what designers ask us for next, not by guesses about what jewelry might be.
Free to start, because gatekeeping kills creativity.
Every other industry got generative AI on its first day. Jewelry shouldn't have to wait, and it shouldn't pay for the privilege. Tashvi's free tier exists because the bar to start designing should be the same bar to start imagining.
Start Building with Tashvi Studio
Tashvi Studio empowers you to unleash your creativity and build anything you can imagine.
Start CreatingTashvi completely transforms design workflows. What used to take days now takes minutes.
