Peer-reviewed work in computer vision, large language model safety, and applied machine learning. This research informs every workshop curriculum and course I deliver.
Three active research threads: medical AI, LLM reliability in enterprise contexts, and predictive analytics for emerging markets.
Deep research in algorithm precision ensures AI safety, data accuracy, and near-zero hallucination in high-stakes diagnostic environments.
A practical playbook for deploying LLMs in regulated industries—finance, healthcare, and legal—with provably lower error rates.
Custom ML models outperform off-the-shelf solutions in emerging market settings—enabling more accurate demand forecasting and resource allocation.
Tools and datasets developed as part of research work, released for the broader ML community.
A modular Python library for preprocessing, augmenting, and evaluating medical image segmentation models. Supports DICOM, NIfTI, and PNG formats.
A benchmarking toolkit for evaluating LLM output consistency, hallucination rate, and instruction-following accuracy across prompt variations.
Curated datasets for training and evaluating ML models on South and Southeast Asian market contexts. Includes retail, agriculture, and financial sectors.
Active and past collaborations with academic institutions, healthcare organizations, and industry research partners.
Scientific ML Lab — computational physics, neural ODEs, differentiable simulation
Enterprise LLM safety — hallucination reduction, output verification, domain adaptation
Medical imaging AI — diagnostic model deployment, multi-site validation, FDA pathway research
Particularly interested in: AI for agriculture, multilingual NLP for South Asian languages, and AI governance frameworks.