We are looking for a Python/MATLAB Programmer for FinTech trading team with an algorithmic trading platform.
- Data transformation, filtering, cleaning and processing (eg computation of various indicators), backtesting.
- Coding and working in MATLAB and Python 3.
- Development and support of data analysis tools for the trading platform.
- Optionally, depending on the candidate’s desire & competence, more sophisticated statistical and machine learning analysis to create strategies (but not required; data processing and coding formulas for existing strategists is sufficient).
- Experience in working with datasets (medium sized, order of several GB), experience in programming, MATLAB and Python (Matlab only also OK).
- Experience with distributed version control (git).
- Reasonable spoken & written English (some team members do not speak Russian).
- Basic understanding of mathematics, statistics, and time series.
Beneficial, but not required:
- Experience in UNIX, scripting (incl. shell scripting) and basic DevOps.
- Basic understanding of markets, trading, tick data etc.
Company needs two broad tasks to be fulfilled, one more junior, one more senior (so they could hire 1 Junior, 1 Senior; or 1 Senior; or 2 Senior candidates).
- Mostly data processing, lots of programming, some mathematics.
- Data cleaning, manipulation, transformation.
- Backtesting infrastructure (take historical price data, take strategy, write code to simulate what strategy would have yielded on given data).
- Mostly data exploration, significant statistics, maybe elements of machine learning, nearly research level.
- Analyse data for patterns and irregularities.
- Formulate trading strategies and backtest them (familiarity with exploration/test/validation advantageous).
- The interface to trading platform is in Python 3 (so Python skills would be advantageous).
- The interface to strategy platform is in MATLAB (so experience in MATLAB required),
- Company is open to expanding to other technologies down the line, but need some results delivered in MATLAB first.
- Вatasets are of medium size (GigaBytes, hundred thousands datapoints).
- Thus, insufficient data for many popular machine learning methods (no neural nets…).
- Goal: classification (supervised learning).
- Preference for old-fashioned linear methods, or
- Very judicious application of some (low dimensional) more modern methods (such as Support Vector Machines, Decision Forest, etc.).
- Small international agile team.
- Flexible work arrangements.
- Beyond our requirements (MATLAB/Python 3 on Linux, git), choose your own tech stack.
- Office near Arbat/Smolensky metro station or remote.
Send your CV to email@example.com