I'm a Quant Researcher at Antarctica Asset Management, building quantitative tools for alternative investments. I spend most of my time writing Python — whether it's probability modelling, risk analysis, or portfolio optimisation.

I hold a BSc Financial Mathematics & Statistics from LSE and an MSc Data Science & Machine Learning from UCL.

🔹 CopulAX — JAX-based probability modelling library with an emphasis on low-dimensional optimization. Univariate, multivariate, and copula distribution fitting with JIT compilation and autodiff support.

🔹 Diana — Locally hosted text-to-speech document converter. Turns PDFs, EPUBs, webpages, and news feeds into MP3 audio using local (Kokoro, Piper) or cloud (OpenAI, ElevenLabs) TTS engines. Includes an AI-powered news aggregator and LLM text cleaning. Built with Streamlit.

🔹 SklarPy — Copula fitting in Python. Implements Gaussian, Student-T, Skewed-T, Generalized Hyperbolic copulas and 100+ univariate distributions via scipy.stats, with integrated parameter fitting and visualisation.

🔹 Languages: Python · SQL

🔹 Libraries: JAX · PyTorch · NumPy · SciPy · statsmodels · pandas · scikit-learn · Streamlit

🔹 AI/ML: Deep Learning · Bayesian ML/DL · Reinforcement Learning · Mamba Models · Transformer Models · Computer Vision

🔹 LinkedIn: Tyler Mitchell