Hi, I'm Deboshree.

I work on machine learning systems and data workflows that help make sense of complex, real-world data.

Machine Learning · Product Analytics · AI Systems

01

About

I started in Electronics Engineering, where I learned to think in systems: how constraints shape decisions, and how components interact. That foundation has shaped how I approach everything I build.

Alongside that, I explored building things in very different contexts, from running a fashion accessories business end-to-end to leading TEDxVITPune as licensee, coordinating a team of 100+ across departments.

After graduating, I spent two years at FinIQ Consulting in a forward-deployed role, working with banking clients across the Middle East and South Asia. I translated business requirements into engineering specs, supported software delivery, and worked closely with stakeholders to ensure adoption post-launch. Here, I was on the team that launched consumer amortising bonds in India!

During my MS in Computer Engineering (Machine Learning) at RIT, I deepened that foundation through work in machine learning, cognitive research, and predictive modeling. I completed my thesis in the predictive mental health space, worked on adversarial ML at Brainlab on an Air Force Research Laboratory-sponsored project, and conducted cognitive research at the SPACE Center studying how deafness affects mathematical reasoning in children.

I now work at Qualitrol, where I build and deploy AI-enabled systems in production: developing ML pipelines for operational data, leading agentic development workflows, and integrating LLMs into engineering tools used across teams.

Deboshree Roy

Across all of it, the motive has been the same: understanding how people interact with systems, and building ones that work for them.

Currently
R&D Engineer @ Qualitrol
Based in
United States