Founding Engineer at Karalabs, YC S26. Co-founder and CEO of Jewellabs. Student @ Cornell University.
I build AI and blockchain infrastructure for the physical economy: the diamond and rare-metal trade, jewelry, and the private equity that moves it. Cornell computer science, blockchain research at Carnegie Mellon, published AI Crop Health work, and product shipped to operators across India, Hong Kong, Dubai, New York, and Los Angeles.
What I am building
Founding engineer at Karalabs, a Y Combinator company (S26). I build computer-vision models that detect and grade diamonds from minimal frame inputs, plus the AI infrastructure and workflow automation that turns them into precision diamond intelligence.
Co-founder and CEO of Jewellabs, AI infrastructure for the global rare-metal, jewelry, and diamond trade, based in Dubai. It ranked in the top 10% of Y Combinator applications. Jewellabs builds fine-tuned models for Fortune 500 houses and global leaders: diamond planning and cutting optimization, AI storefronts, smart CRM, compliance and DDS automation, and diamond data licensing. Clients include DiamondPlugLA, PalakDiam, and Indriya across India, Hong Kong, Dubai, New York, and Los Angeles.
CIO at KOOV, an AI company working with major names in Indian finance and industry, including JM Financial in private equity, Go Cheese, and BDR Pharma, plus construction and metals clients across the United States. Backed by angels from G-Fellows and Cornell Commons.
A nonprofit that pairs school districts across the United States with student mentors, supporting students on IEPs and 504 plans with time management and academic skills. Recognized on the Forbes 30 Under 30 list. I build outreach systems as CXO.
An AI social-media analytics app for creators that scores and optimizes post performance from engagement data. I led product strategy, growth, and feature direction.
Research
Researching on-chain provenance for physical stones at Carnegie Mellon: scan each diamond's physical fingerprint onto a blockchain so it cannot be cloned and its history cannot be backdated. The authentication layer the trade has never had, and the same trade Jewellabs already serves.
A convolutional neural network that diagnoses crop disease from images, which I built and trained with GPU acceleration and deployed at a local farm to lift yield through early detection. Published on IEEE's preprint platform with over 2,000 views and 100 downloads, and recognized with Technical Excellence awards from CREST Gold and IRIS.
Experience
At a Los Angeles diamond startup, I consulted for international stakeholders, using deep learning on historical sales data to tighten lab-grown diamond supply chains, predict design trends, and price B2B orders across Hong Kong, UAE, and LA, securing over 10 high-value orders.
Sold natural and lab-grown diamonds, including CVD and HPHT goods, and priced parcels for firms across Hong Kong, Mumbai, Dubai, Bangkok, and Los Angeles. The trade Jewellabs now builds for, learned at the desk.
Analyzed private-equity and quantitative-trading data in Python and R at the Haas School of Business, studying how geopolitical shifts and fintech firms move stock volatility and trading behavior.
Ran digital marketing and social for a Hong Kong watch company, across PR, ad campaigns, and both front-end and back-end web work to lift reach and conversion.
Ran cold-outreach client acquisition, from lead generation to automated follow-up, contributing to closed deals with Genentech and Porsche.
Built full-stack AI in Python with Django, Flask, NumPy, and Pandas: face-following drone navigation, CIFAR-10 image recognition on 60,000 images, and pathfinding for robotics, across Python, C, and Java. Around 30 programs in total.
Earlier building
Founded a tech-enabled environmental policy initiative in peri-urban India. Built machine-learning diagnostics to estimate regional air quality and model mitigation for pollution hotspots, plus digital hubs linking stakeholders with policymakers.
A cross-platform app that recognizes vehicles, predicts emissions, and calculates the tree-planting offset for carbon neutrality, with a carbon-credit score for future trading. A donation funnel I built raised over $700.
About
The thread through everything is the same: take intelligence that works in a notebook and put it to work where money and materials actually move.
I study AI at Cornell. I have sold diamonds across Hong Kong, Dubai, and Los Angeles, priced parcels for the trade, analyzed private-equity and trading data at Berkeley Haas, published computer-vision research that runs on a working farm, and shipped an app to the App Store.
Now I build AI and blockchain infrastructure for the trades I grew up around. That means Jewellabs for the rare-metal and diamond world, founding engineering at Karalabs, information strategy at KOOV, and research at Carnegie Mellon on putting a diamond's physical fingerprint on-chain so it can never be cloned or backdated. I like problems where the model has to survive contact with a real operator, a real supply chain, and a real invoice.
Away from the screen: cricket, football, chess, and watches.
Recognition and education
Coursework and certifications
Capabilities
Contact
The fastest way to reach me is a 15-minute call or an email. I reply to founders, investors, and operators.
CONNECT