Adithya Kethu

At Origin Global, we design AI agents & compliance software to run back-office ops for SME physical commodity traders & manufacturers. Origin Trade supports trade desks across contracts, vessels, compliance, demurrage, and commissions, with over $4M in trade flow processed. Origin Run supports factories across planning, production, inventory, QC, maintenance, dispatch, finance, and WhatsApp operations, and is currently in pilot with a manufacturer representing roughly $60M in enterprise value.

I love bringing ideas to life. Digipuja helps younger generations stay rooted in their traditions, with paying users and over 200K visits in its first month. RantBox.app continues to be used by people discovering it organically, CramCram.io is a full AI-powered learning suite used by students at top universities, and FYOP was my answer to campus isolation.

At 20, I grew from an intern to Director of Technology at a DentalTech company that raised $500K, after which I exited to build full time. I led the technical strategy and shipped Gnathic, the world's first non-invasive, photo-based jaw malocclusion detector, and Pacified®, now trusted by 2000+ parents in the US.

Co-authored a peer-reviewed paper in Children (MDPI) on using data science for optimal pacifier sizing, and a second Children (MDPI) paper on mandibular growth, airway health, and pacifier shield biomechanics. That research helped validate the science behind Gnathic and Pacified®. Also presented ML research at the World Sleep Conference and Harvard's Global AI Dental Symposium. Separately, Origin Global was selected as one of 50 startups worldwide to pitch at MIT Imagination in Action.

At Wayfair, I worked on Search & Recommendations, building backend systems for Big Data, ML, and SEO that drove over $10M in revenue impact across millions of users. Shipped a no-code ML tool that made it way faster for teams to deploy models. I left to build systems for high-stakes operational workflows in trade and manufacturing through Origin Trade and Origin Run.

As Tech Lead at BUILD UMass, I led a team of engineers and a PM to build an OCR inventory tool for a local business. At HackUMass, I engineered internal tools and mentored hackers. In the lab, I worked with Prof. Marco Serafini at the UMass DREAM Lab to optimize graph sampling for 'NextDoor' using deep learning, and analyzed human-subject trends in CS research with Prof. Ali Sarvghad at the UMass HCI-VS lab.