- 18+ years in global banking technology (ANZ, Goldman Sachs, JPMorgan, Bank of America, Barclays)
- Currently completing my M.S. in Information Science (Machine Learning track) at the University of Arizona campus (graduating Dec 2025)
- Hands-on with Machine Learning (LLMs, deep learning, explainability), CUDA programming, microservices, large-scale distributed systems, and ultra-low-latency platforms for finance
-
Ultra Low Latency L2 Order Book Prediction
LSTM/Transformer models on financial order book data for real-time, ultra-low-latency price movement prediction. -
Interpretable RAG Fact-Checking
Qwen2, GPT, Mistral LLMs + FEVER dataset, PEFT/QLoRA + DPO + explainability. -
Gallager-B CUDA Optimization
Parallel batched decoding, CUDA streams, single-block kernels. -
Retinal Segmentation
End-to-end ophthalmic imaging pipeline: dataset curation (frame extraction & annotation with Supervisely), multi-label segmentation** of retinal structures, and training deep CNN/Transformer models for segmentation and downstream disease classification tasks. -
Smartcard Ticketing Microservices (NTT Data Interview Demo, 2020)
Spring Boot + Kafka + Spark + Docker; ~7K events/sec prototype throughput.
This GitHub includes academic, prototype, and interview projects that I am free to share.
My professional work over 18+ years in banking (low-latency trading, risk/analytics platforms, large-scale distributed systems) is confidential, proprietary and cannot be published here.