When people hear I'm a Data Science student who also runs a jewelry business, the reaction is usually confusion. What does machine learning have to do with gold necklaces? More than you'd think.
The Pattern Recognition Connection
Data Science is fundamentally about finding patterns in noise. So is running a century-old jewelry shop. My family has been making decisions about gold inventory, customer preferences, and seasonal demand for generations — all based on intuition built over decades. Data Science just gives us better tools for the same problem.
At UCSD, CSE 258 (Recommender Systems) taught me how to model customer preferences mathematically. I immediately saw applications: which jewelry styles sell best during wedding season? Which customers are likely to come back for anniversary purchases? What's the optimal gold inventory level given price volatility?
Graph Theory Meets Jewelry Design
My fascination with Graph Theory started in DSC 80 and deepened through independent study. Graphs model relationships — and jewelry design is all about relationships between elements. The vertices and edges of a necklace pattern, the symmetry groups of a pendant design, the network of suppliers and craftsmen. Even the 3D icosahedron visualization on my portfolio is a nod to this connection — a graph-theory structure that doubles as a geometric jewel.
What Entrepreneurship Teaches That Classes Don't
Running Gondilal Saraf taught me that the best model is useless if you can't explain it to someone who doesn't know what a model is. My uncle doesn't care about Prisma migrations or SDXL inpainting — he cares that the website loads fast on his phone and that uploading a product photo takes less than 30 seconds.
This constraint made me a better engineer. I obsess over performance (our customers are often on slow connections in rural UP), I write interfaces that are obvious (the admin dashboard uses Hindi labels), and I never add complexity that doesn't serve a real user need.
The Minor in Entrepreneurship
UCSD's Entrepreneurship & Innovation minor gave me frameworks for the intuitions I was already developing. Lean startup methodology, customer discovery, financial modeling — these aren't abstract concepts when you have a real business to apply them to. Every assignment became a chance to improve Gondilal Saraf.
What's Next
I graduate in June 2027 with a BS in Data Science and a minor in Entrepreneurship & Innovation. I'm looking for roles where I can combine technical depth with business impact — whether that's ML engineering, data science, or building products. If you're working on something interesting, I'd love to talk.
The best part of my unusual path is that I never have to choose between the analytical and the creative, between the technical and the entrepreneurial. They're the same thing, viewed from different angles — just like a well-cut diamond.
Written by
Arnav Goel
Data Science senior at UC San Diego, shipping ML systems and a 150-year-old family jewelry business.