Improve market making performance with HFTFramework, a research-backed tool for founders in high-frequency trading.
advancedโฑ 1-2 hours๐ต Free (self-hosted)
303 stars61 forksJupyter NotebookQuality 8/10Updated 7/18/2026100% free ยท open source
What it is
Use HFTFramework to create algorithms that make markets more efficiently.
What you can make with it
Automations like optimizing market-making performance for your high-frequency trading system.
How it helps
HFTFramework improves market making performance with research-backed strategies, making it a valuable tool for founders in high-frequency trading.
Real use case example
"A founder wants to improve the performance of their high-frequency trading system. They use HFTFramework to implement a reinforcement learning approach to the Avellaneda-Stoikov market-making algorithm. Within a few hours, they've optimized their system to execute trades more efficiently."
If you're new
You should pick this up if you're new to high-frequency trading and want to learn from the research.
If you're senior
A senior engineer/professional would use this for its research-backed strategies and high-frequency trading expertise.
Common confusion cleared up
HFTFramework is primarily designed for high-frequency trading, making it not suitable for general machine learning tasks.
Best inside these AI tools
Claude DesktopClaude Code
Pairs with
Claude API
Why we list it on WorkflowStacks: This is an open-source, research-backed tool for high-frequency trading that a marketplace of AI tools includes due to its unique, specialized focus.
What it does
HFTFramework improves market making performance using a reinforcement learning approach to optimize the Avellaneda-Stoikov market-making algorithm
Install / run
git clone https://github.com/javifalces/HFTFramework && cd HFTFramework
When to use it
โขWhen you need to optimize market making strategies for high-frequency trading
โขWhen researching reinforcement learning applications in finance
โขWhen looking to improve the performance of the Avellaneda-Stoikov algorithm
Quick start
1Run the Jupyter Notebook by executing `jupyter notebook` in the terminal
2Open the `HFTFramework.ipynb` notebook and follow the instructions
3Modify the `config.py` file to set up your market making environment
4Run the `train.py` script to start training the reinforcement learning model
5Evaluate the performance of the optimized market making strategy using the `evaluate.py` script
Ready-to-paste prompt
Run the `python train.py --env=stock_exchange --agent=rl_agent` command to start training the reinforcement learning model for a stock exchange environment
Heads up: Make sure you have the required Python libraries installed, including `numpy`, `pandas`, and `gym`, and that your Jupyter Notebook is configured to run the necessary dependencies
Saves to your device
Topics
algorithmic
avellaneda
avellaneda-stoikov
deep-learning
hft
high-frequency-trading
market-maker
market-making
market-making-bot
reinforcement-learning
stoikov
trading
trading-bot
trading-strategies
What's inside โ free to inspect
No purchase needed
Read the entire source before you build โ unlike paid marketplaces that hide it behind a buy button.