• Algorithm design
  • Simulator
  • Execution framework for your data/order execution provider
  • Julia for rapid prototyping
  • Time-series analysis

Neural networks

  • supervised and un-supervised methods
  • Multilayer feed forward neural networks
  • Deep auto encoders for nonlinear compression/mapping
  • modern training methods
  • massively parallel execution
  • feature extraction

High performance computing on your hardware or Amazon Elastic Cloud

  • MPI-3 on reliable clusters ie: Elastic Cloud2 dedicated instance
  • openMP for efficient parallelization
  • c++11 threads/posix threads for fine control
  • CUDA on NVIDIA GRID/TESLA GPGPU
  • protocol buffers/thrift
  • HDF5 scientific datastore for large data