About

I build production AI systems for complex industrial operations. My work combines machine learning, product engineering, and customer delivery, taking solutions from an initially ambiguous operating problem through modelling, validation, deployment, and production support.

Currently, I am a Forward Deployed ML Engineer at Applied Computing, working on decision systems for refineries and energy operations. I collaborate directly with process engineers and operators to turn plant data, engineering knowledge, and operational constraints into forecasting, anomaly detection, and optimisation capabilities that fit real workflows.

My background spans computational geophysics, industrial research, computer vision, generative AI, and production ML. That combination helps me move between scientific problems, software systems, and the practical realities of deploying technology for customers.

Before that, at Jiffy, I built the AI image enhancement pipeline powering a $100M+ DTF printing business, architected Smart Redraw (a multi-agent AI editing system), and developed end-to-end ML solutions spanning LLM fine-tuning, custom model deployment on RunPod/Replicate and LLM-powered automation.

Previously at Flixstock, I integrated the HuggingFace ecosystem as core ML infrastructure and built Stable Diffusion pipelines for background generation, identity transfer with pose control and fashion video generation.

Unlike many in this field, my core background is not 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 three years. I believe this diverse background helps me approach problems from unique perspectives.

I love connecting and collaborating with new people. Please don’t hesitate to reach out! 😊

Professional Experience

Applied Computing

May 2026 to Present

ML Engineer (Forward Deployed)

Build and deliver production AI systems for refinery and energy operations, working directly with process engineers and operators from problem framing and industrial data integration through model validation, deployment, and production support. Translate plant constraints and customer feedback into forecasting, anomaly detection, and optimisation capabilities that industrial teams can understand and trust.

Jiffy

Jan 2024 - Mar 2026

Senior Machine Learning Engineer

Built and maintained the core image enhancement pipeline, Smart Redraw (a multi-agent AI editing system) and ML solutions spanning fine-tuning, model deployment and LLM-powered automation.

Flixstock

Aug 2022 - Dec 2023

Software Development Engineer III

Built Stable Diffusion pipelines for background generation, identity transfer and video generation.

Shell

Oct 2019 - Jul 2022

Research Geophysicist

Built DL-based denoising models for seismic data, worked on uncertainty quantification and DL-based Hessian operators for wave-equation optimization.

TU Delft

Aug 2015 - Aug 2019

PhD Researcher

Developed JMI-res for high-resolution subsurface imaging and created a DL-based super-resolution approach for seismic spatial aliasing removal.