YouTube

YouTube

- End-to-end delivery of complex mathematical / statistical models from ideation to production ready high-performance quality code- Designed and built scalable, robust parallel analytics computing system serving 100k+ fixed income securities per day- In-depth knowledge on pandas, numpy, core PythonPython Quantitative Analytics Infrastructure- Designed analytic production pipeline concept and implemented monitoring server with logging engine and testing framework for smooth production rollout- Developed OOP template of quant models and designed well-structured APIs saving significant development time for the team- Built core analytic engine consists of Python wrapper for third party C library for foreign language function interface (cffi), and numpy memory view- Prototyped fixed income analytic models including treasury future and option, bankloan in Python serving as team reference code- Applied map-reduce to service to parallelize quant model execution. Speed gains up to 100+ times- Tested distributed computing frameworks (Spark, AWS, dask) and GPU acceleration of Monte Carlo simulation (numba) for further speed-up- Worked with IT to integrate the project to Github/Jenkins version control pipeline, Autosys job scheduler, and signed off dev / qa / prod Linux environment setups- Internal instructor of the team to share in-depth knowledge for effective and efficient Python coding- Contributed team’s Python code development/refactoring through Python readability and knowledge- On-call support on production failureMachine learning Use Case Analysis- Feature selection/engineering, dealt with overfitting issues, interpreted families of machine learning models- Built Python test pipeline from data wrangling for different ML APIs to report generationFund Performance Attribution- Extensive Data wrangling / analysis (SQL) covering all layers of internal data, from trade tickets to fund accounting.- Provided detail oriented quantitative analysis

Extensive clean-up and building different statistical data sanity checks on internal state-of-world market data database. Investigate data anomalies and proactively work with different data vendors to identify the issue and fix them. Lots of manual team process automation using SQL, SAS, VBA, Access.

- Developed brief cash-flow forecasting program using VBA- Customized CMPR's OPUS-II portfolio optimization engine to meet Financial Supervisory Service's new risk-based capital ratio regulation, using C++- Developed money management system which is coherent with existing risk management system in KHFC, mainly using PL/SQL- Quality analysis

- Rotated different teams - emergency communications center, fire & rescue, paramedic.- Worked under fast-paced, time sensitive, heavy pressure environment