Our Best AI-Powered Crypto Intelligence Tool Yet
The GNY Range Report uses AI to tune out the noise and emotion of crypto trading, and just deliver the information you need to make more data driven decisions. The delivery of these insights just got even better with the introduction of our custom Large Language Models.
When GNY launched the Range Report two years ago, we focused on developing the most accurate predictions for Bitcoin. We are committed to developing technologies at the intersection of blockchain and machine learning, and predicting BTC price volatility seemed like the best place to start. Building proprietary LSTM machine learning models allowed us to deliver weekly predictions which evolved into daily predictions — but the market is more than just Bitcoin so we started to grow the report.
As the GNY Range Report expanded to ETH and then 12 top tokens we heard consistent feedback from traders that loved the platform but weren’t exactly sure what to do with the information. A single page of indicators and volatility predictions didn’t quickly communicate the changes in trend and volatility they needed to make a decision or decide where to invest more research time and attention. The main thing users asked for was clearer signals, and more distilled information.
Most crypto traders are part-time and don’t consider themselves professional traders. Often the impulse to check in on an asset requires more than simply checking the current price. You want to understand how the price is changing, what is the momentum doing, and what about the indicators that you have found to be most helpful trading that token? Our goal was to provide a summary that would provide users an up-to-date roadmap of the token’s current behavior and to the most relevant charts so they can make informed decisions on whether to buy, sell, or trade the asset.
We needed to develop a way to make the insights more accessible and for users to follow the invisible trends woven through the report to develop their own understanding to bring to trading decisions. The introduction of Large Language Models like OpenAI’s ChatGPT and Meta’s LLaMa 2 offered the solution. The goal of these new tools was to provide users summaries of the findings of the Range Report so that they can have a framework and guidance to better connect how all the figures in the report point towards signals a trader can use to make more objective decisions.
Like other successful integrations of LLMs into businesses, our approach to this challenge involved centering the execution around our value proposition and feeding our machine learning models our proprietary data. The results are impressive and in the months that we have been testing in our development site, we have been counting down the days until we could share this tool with our community.
We see three clear scenarios how these Chart and Token Readouts can simplify your trading routine.
- Helping traders notice significant changes in trends
- Highlighting specific changes in an indicator’s behavior for a specific token
- Providing a daily guided review of the charts to improve a trader’s chart literacy
These characteristics combined, elevate the knowledge of the trader considerably while saving them time. We hope that you take the time to explore the new tool and see what it can bring to your trading approach.
Although the results we’ve obtained from the implementation of LLMs have been impressive, we’ve also understood their weaknesses. They do make mistakes. Although we’ve implemented layers to check their output, it will be impossible to catch them all. We will continue to make the process more robust. Although we believe that the Readouts will help traders considerably, always check and verify the information.
Our next innovations will focus on the building of DataNFTs to bring transparency and auditability to our reports and eventually to ML transactions and products of all kinds. Subscribe to stay tuned for more updates.