Centurion Capital is a quantitative investment management company researching in global financial markets, dedicated to producing exceptional returns by strictly adhering to mathematical and statistical methods.
Our researchers are responsible for creating the programs that generate investment returns. They are given the freedom to explore and think differently. A collaborative environment means that individual expertise is shared to strengthen the aggregate result.
- Mastery of Python 3
- Mastery of Django framework
- Good knowledge of Django REST Framework
- Expertise with relational databases (preferably PostgreSQL)
- Passion for writing maintainable tests (unit, functional, end-to-end, mocks)
- Solid understanding of design patterns and OOP
- Expertise in VCS: Git and/or Mercurial
- Accuracy in writing code
- Experience in Agile development
- Ability to work to tight deadlines
- Passion for building great software
Key activities will include:
- Building scalable analytical codebase for data preprocessing and models training/validation;
- Training machine learning models suitable for retail business data;
- Doing A/B testing and econometric evaluation of results;
- Creating ideas for new data and modeling tasks;
- Mentoring other team members on model building and data management.
- Good understanding of standard models for prediction. Linear models with variables selection, boosting, neural networks;
- Good understanding of bias/variance tradeoff, nested cross validation, regularization techniques in different model settings;
- Knowledge A/B testing, econometrics estimation and experimental design is required;
- Ability to quickly read papers and learn models from previously unknown areas of machine learning.
- Experience with standard model fitting packages in Python (scikit-learn, pytorch/tensorflow, xgboost). Density estimation techniques;
- Experience with econometric estimation is preferred;
- At least 3 years of experience;
- Experience with SQL/NoSQL databases;
- Good communication skills.
- PhD or Masters in Mathematics, Physics, Finance, Statistics, Computer Science or a related field;
- Kaggle participation would be a plus.