Steven Signal: AI Trading Innovation

Empowering investors with real-time insights: Meet Steven Signal, your AI trading companion for Telegram.

Overview

Mad Devs created a pet project Steven Signal. This AI-powered trading bot recognizes trading patterns from popular stock markets and sends real-time updates to the popular messaging app Telegram.

The goal of the ongoing project is to equip traders with expert trading analytics in a user-friendly, and convenient format, explaining to them what kind of data it is and what they can signal. Steven Signal prompts users about market events, helping them to save time and increase their investments.

Please note that the signals generated by the bot are not financial recommendations. Use the information at your own risk and exercise caution. Do not blindly follow the bot's recommendations. It's important to validate the signals with your own trading or investment strategy and analysis.

With Steven Signal's user-centered interface and real-time updates, traders can make well-informed, data-backed decisions. Whether one is a newbie or a seasoned investor, this powerful tool empowers every user to navigate the markets efficiently and seize profitable opportunities.

How did the project come to life?

We have noticed a huge market of minnows, non-professional crypto traders, and rookies, enthusiastic and aspiring traders, who are just in the market with very little capital and the greatest desire to earn money. They have a thing in common: their goal is to quick profits from trades and reinvest them into personal items, rather than future exchanges.

We conducted research and realized that only a small percentage of traders earn money from sales. Some statistics claim that just 28% of investors make profits.

There were 2 identified reasons for that:

Minnows and rookies haven’t studied the market and, hence, don’t have a sense of trading intuition and the skill of pattern recognition.
They get manipulated by whales, seasoned crypto traders with a larger capital and influence on the price.

Our team seized the opportunity and leveraged our technical expertise to develop an AI-powered bot that recognizes stock patterns and helps beginners in trading.

Making everyone equal

The mission of Steven Signal is to fill in the niche of predicting trading patterns on crypto exchanges and provide all traders with equal opportunity in their professional activities. Being a pet project in the beginning, the Steven Signal has gradually grown and acquired advanced functionality, such as the addition of the most typical stock indicators in the analysis.

I personally benefit from Steven Signal’s feature for identifying candlestick patterns and prompts insights into whether it is worth staying in a trade or not. The bot itself determines levels, gap zones, and patterns, and also shows volume profiles. It is difficult to track the behavior of dozens of trading pairs at the same time, but the Steven Signal makes it happen by focusing on informative signals and assisting with trading decisions. I am glad to notice that the bot is constantly evolving and new features are getting added. I appreciate that Steven Signal does not trade for me, does not force transactions on me, does not shout “Buy or sell right now,” but rather provides useful information and allows me to see the full picture, without being by a computer 24/7.

Vitaly Skripov, Trader

Project hypothesis

The later traders enter into trades after the whales have manipulated the market, the more money minnows lose and the more the whales earn. If minnows receive a timely signal that a trading pair is starting to go up or down, they would have time to ride the whale’s wave to make money going long and short. Therefore, if minnows have the tools to recognize the trend early, they can make profits.

Challenges & Solutions

Since the price of cryptocurrencies can depend on others and there are approximately 30,000 currency pairs, for Steven Signal we had certain requirements for infrastructure and algorithms that identify patterns.

As in any project, we encountered some challenges in developing the idea. Here is an example of how we solved them.

Challenge

The requirements complicate the development cycle of pattern detection algorithms since it does not allow the use of neural networks and computer vision. The more GPUs we needed, the bigger the budget for infrastructure would be.

At first, the Steven Signal was coded in Python and developed as a monolithic solution that cannot solve different tasks at once. The architectural solutions and libraries used in the sandbox could not process the data at the required volume and in a compressed time frame in a sprawling project.

Solution

We amped up project architecture and switched to Domain Driven Development (DDD) principles, as it allowed us to test and optimize the project more flexibly. Instead of multithreading, we introduced the Asyncio, which is ideal for working with multithreading. In addition, we chose SQLAlchemy over PonyORM to ensure stable work with databases.

Key results

Stable performance

Now Steven Signal operates autonomously 24/7, 365 days a year. Signals are sent exactly at the set time, and the bot hasn’t encountered any failure.

Flexibility

The new architecture allows users to easily integrate signals about stock indicators, such as EMA, SMA, and RSI, and add third-party behavioral patterns, or new data processing layers.

Machine learning

It is planned to add a machine learning library capable of analyzing its own candlestick and figure patterns to predict trend movements.

What’s next

Steven Signal is an ongoing project, and our team has an ambitious long-term strategy for its development. Currently, the bot is at this stage when success depends on the right marketing strategy. We actively work to raise attention around the product, use cutting-edge online marketing tools, and also build a client base by surveying potential users and collecting feedback.

We plan to add optional trading alerts and improve trading pattern recognition algorithms, prioritizing the quality and functionality of the product. Moreover, we aim to launch a similar bot for Discord and open it to a wider audience by entering the international market and using innovative machine-learning solutions and professional trading strategies.

Conclusions

Despite the early stage of the project, Steven Signal has proved its long-term prospects for growth and expansion. At the moment, the project team has already implemented a significant part of the trading bot’s functionality and created a landing page to establish first contact with potential users.

Taking into account the efforts invested in development and long-term plans, we are certain that this project has the potential for successful development in the field of trading. Decisive steps towards improving the product and attracting customers will contribute to its growth and strong positioning in the market.