I am an applied machine learning engineer passionate about building AI-driven products to solve real-world problems. My motto is “Applied AI should ALWAYS be product-driven”.
Currently, I am working as a Senior Machine Learning Engineer at Jiffy, where I focus on building in-house Generative AI-based image processing features and leveraging Large Language Models (LLMs) and Vision-Language Models (VLMs) for search and recommendation tools. Of late, we as a team focussed a lot on generating low-cost detailed structured metadata for the different assets (fonts, designs, products).
Previously at Flixstock, I predominantly developed vision Generative AI solutions, with a focus on building Virtual Try-On and Background Generation solutions emphasizing identity and pose control. We constructed product pipelines around StyleGANs, Vision Transformers, Stable Diffusion, and its variants such as ControlNets, T2I-Adapters, and Dreambooth.
Unlike many in this field, my core background isn’t in computer science. My graduate and Ph.D. research experience is in computational geophysics, and I worked for Shell as a research geophysicist for about 3 years before transitioning to applied deep learning. I believe this diverse background helps me approach problems from unique perspectives.
I love connecting, chatting, and collaborating with new people. Whether you’re a fellow professional, a new graduate, or someone interested in discussing AI-powered products, please don’t hesitate to reach out! 😊
You can also book a meeting with me using the Calendly link.
Professional Experience
Jiffy
Senior Maching Learning Engineer
- Working on building Generative-AI image processing features and leveraging LLMs/VLMs for search and recommendations.
Flixstock
Software Development Engineer III
- Developed vision-based Generative-AI solutions for Virtual Try-On and Background Generation.
Shell
Research Geophysicist
- Focused on uncertainty quantification and accelerating denoising of subsurface realizations/data.
TU Delft
PhD Researcher
- Worked on a new imaging approach, JMI-res, for high-resolution subsurface physical properties.