Quantitative Researcher - ML Trading Systems

Claire Niu

*************@*****.*** 929-***-**** http://www.linkedin.com/in/claire-niu

Education

UCLA Anderson School of Management Los Angeles

Master of Financial Engineering Sep.2023-Dec.2024

Curriculums: Asset Pricing & Derivatives, Fixed Income, Stochastic Calculus, Numerical Methods, Machine Learning, Market Microstructure

& Algo Trading, Quantitative Asset Management, Financial Computing (Python/C++), Risk & Credit Modeling The Chinese University of Hong Kong Hong Kong

Bachelor in Financial Mathematics (Dean’s List 2020-21, Dean’s List 2019-20) Sep. 2018-Jun.2022 Skills

Programming: Python, R, C++, SQL, Stata, MATLAB

ML/Quant Libraries: Pandas, NumPy, scikit-learn, TensorFlow, PyTorch, HuggingFace GenAI & LLM: LangChain, FAISS, RAG/CRAG pipelines, Ollama, Gemini API, Agentic Frameworks Data & Tools: FactSet, Bloomberg Terminal, Git

Professional Experience

Google DeepMind-Instalily Hackthon Project San Francisco InstaWorker Feb.2026-Jun.2026

Built RAG/CRAG pipeline with FAISS for warehouse task routing; corrective retrieval fallback eliminates hallucination on out-of-distribution queries.

Designed three-layer agentic fallback engine (Gemini Ollama phi3 rule-based) with five-state task state machine for graceful degradation under edge connectivity loss.

Anderson FIS Fund (UCLA) Los Angeles

Quantitative Researcher (Multi-Asset Systems) Feb. 2025-May.2026 Built a multi-asset forecasting and analytics system combining yield-curve factors, macro indicators, equity/FX signals, and cross-asset correlations. Improved predictive accuracy 15–22% and reduced RMSE 17.5% in curve-factor modeling. Developed curve-shape and rate-directional indicators derived from factor dynamics and macro interactions, used for identifying directional and RV structures in rates markets.

Constructed cointegration & Kalman-filter spread systems across equities and credit curves, reducing mean-reversion half-life 28% and delivering out-of-sample Sharpe 1.0–1.3.

Built a backtesting & scenario-evaluation framework analyzing signal decay, regime shifts, robustness and execution sensitivity, reducing research iteration time 39%.

Strat-AI Los Angeles

Machine Learning Researcher Mar.2024-Aug.2024

Developed LGBM margin-risk models by incorporating richer market features (rate shocks, vol regime, liquidity spreads, exposure dynamics) and tuning regularization / class-weighting / sampling schemes to improve scenario-ranking 17% precision and reduce tail-risk false negatives.

Built curve-factor forecasting components that enhanced directional and curve-shape signal stability under macro-regime changes and improved downstream portfolio-risk evaluation.

Designed a reinforcement-learning-based dynamic allocation engine, optimizing reward shaping and state features to improve Sharpe 0.32– 0.45 under stress environments and productionized

Singular Researcher Los Angeles

Quantitative Researcher (AFP) Oct. 2024-Feb.2025

Haitong Securities Shenzhen

Bond Execution Intern May. 2021-Sep. 2021

Built fixed-income analytics infrastructure for rates, credit and MBS/CBS, including curve construction, OAS, DV01/KRD decomposition, and spread-risk metrics; integrated outputs into daily pricing & risk workflows. Designed automated curve/spread refresh pipelines, reducing data latency 40% and improving intraday exposure visibility 22% across FI books.

Developed scenario & stress-testing modules for rate shifts, credit migration and refinancing shocks, enabling traders to simulate PnL and risk under market moves.

Produced issuer-level credit diagnostics and MBS pool analytics that accelerated curve-shape and sector RV screening 30%, supporting directional and RV trade evaluation.

Award

2025 International Quant Challenge - Gold Medal Jun. 2025 Kaggle Silver Medal:"Stable Diffusion - Image to Prompts" Jun. 2024

Contact this candidate