Implementation of Machine Learning in MATLAB from predesigned architecture

I am looking to have a Machine Learning time-series training and deployment code written in MATLAB.
Specifically, the code is designed to take forex price and specified indicators and train a combined Reinforcement Learning and CNN approach. The architecture of the approach is already drawn up and only requires implementation. There should be two modules created, one that trains and tests the RL agent, and another that can be used to deploy it, taking in input data and making predictions. The model code should also be setup in a clear and structured way that can easily be adjusted later by a MATLAB programmer without difficulty.

The following is available:
- Detailed description of the model architecture
- Sample input data for training and testing

The skills and experience required for this job are:
- Expertise in MATLAB deep learning and reinforcement learning
- Some stock/forex knowledge can help, but is not required
- Ability to offer further suggestions how the model can be further fine-tuned is also desired