top of page

Reinforcement Learning Engineer

About Us

Block by Block Capital was founded in 2017 and has grown to become a technology-driven, diversified principal trading firm with our flagship product being an algo based hedge fund. We also trade our own capital at our own risk, across a broad range of asset classes, instruments and strategies, in financial markets around the world. Within the past year, we’ve turned over ten-figure volume in crypto-to-fiat arbitrage and algo trading and continue to prioritize the growth of our fund and automated trading footprint.

Position Description

We are seeking a senior reinforcement learning developer who can drive the architecture design and spearhead the hands-on implementation. You must have past experience in deep reinforcement learning and be able to lead team discussions and implementation planning for the appropriate ML frameworks, underlying learning models (or building custom ones), and structuring rewards policies to optimize the performance of our model's potential.

Relevant Skills

  • Reinforcement learning

  • GPU / Tensor / Python

  • Trading domain knowledge

  • Model development

  • Algorithm optimization

  • Rewards policy design

  • ​Parallel agent learning frameworks

Primary Candidate Considerations

  1. Core Skills in Python and ML/RL:

    • Python: Strong proficiency in Python for both general programming and specialized ML/RL development

    • Reinforcement Learning (RL): Deep understanding of RL algorithms, including DQN, Proximal Policy Optimization (PPO), and Advantage Actor-Critic (A2C)

    • Machine Learning (ML): Practical experience in applying ML logic

    • Deep Q-Networks (DQN): Proven ability to implement and optimize DQN agents for financial market simulations

  2. Framework and Libraries Knowledge:

    • TensorFlow: Expertise in frameworks for building and training deep learning models, particularly for RL

    • Gym: Familiarity with OpenAI Gym for building, training, and deploying RL agents

    • NumPy/Pandas/Scikit-learn: Strong knowledge of data manipulation and scientific computing libraries, especially for pre-processing and backtesting trading strategies​

  3. Parallelization and Performance Optimization:

    • Multi-agent RL Systems: Experience designing systems with multiple agents interacting in parallel

    • CUDA/GPUs: Ability to utilize GPUs for training deep RL models, optimizing performance for real-time trading environments

    • Distributed Computing: Knowledge of parallel computing architectures (e.g., distributed systems, multi-threading, or cloud-based computing solutions like AWS or Azure for model training)​

We Offer

  • Hands-on experience with technical trading, financial markets and hedge-fund operations, while working directly with the company founders and broader tech team

  • Opportunity to learn about blockchain and cryptocurrency industry

  • Flexible working hours and time off in a small, tight-knit family company

Location

  • Remote

Rockstar Talent Application

Thanks for reaching out. Our investment team will get back to you shortly.

bottom of page