Don’t invest your emotions in the stock market.

Invest in a bot. 🫶

+37% is what our bot has earned for customers since its commercial launch in September ’22, with a potential annual profit of up to +69%

Check the results

Two strategies for different stock market scenarios – a bear and a bull market. BTC/USDT on Binance. Check which strategy is best for you. Both are doing very well. Conservative more for bears 🐻, Dynamic more for bulls 🐂.

Dynamic Strategy

The Dynamic Strategy, due to its rather aggressive assumptions, works best during an upward market. That’s why we present the results for 2021, which was clearly a bull market.
CLICK HERE to take a look and download the detailed backtest data for your analysis.


Conservative Strategy

Conservative strategy is ideal during a bull market and sideways trend. Thus, we present the results for 2022, which was the best for this strategy, as shown by the results achieved.
CLICK HERE to take a look and download the detailed backtest data for your analysis.


How does it work?

Human emotions are bad advisors, especially in financial markets. Major funds are taking advantage of the benefits of behavioral finance by measuring sentiment. They know that action follows emotions.

Our SentiBot reads investor sentiment. A unique crypto trading bot, which uses market emotions to determine what operation (SELL/BUY/HOLD) is best for a given moment. Therefore, it is one step ahead of the market. Every hour, continuously for more than two years, it makes decisions based on the emotions measured in mentions.

How does it do it?

A total of 5 years of work by top scientists, 8 basic emotions, dozens more emotions, 14 million tweets and 15 million candles to train an artificial intelligence model to forecast the price of BTC. That’s how our power couple – Catherine the Machine & AI-fi – was created. Heart and Reason.

From now on, you can also take advantage of this. Choose one of three subscriptions and regain power over your emotions. Use them better, it’s a shame to waste them on the stock market.

Catherine the Machine

Our first AI, the NLP engine. She is the one who reads each of the eight emotions, sentiment, and arousal, tirelessly doing so on thousands of arbitrarily long texts, simultaneously, in a dozen languages.

She can read the context of an utterance, understand emoticons, acronyms, overtones and temperament. By understanding context, Catherine can also categorize texts.



Our second AI, the forecasting engine. AI-fi* – an open-batch prototyping module.

For now, combining emotional data (based on Catherine the Machine’s emotions) from mentions shared on social media with market prices of a given asset allows AI-fi to find correlation and predict the rate.

Open ended, modular, ready for any market, language, or country.


Frequently Asked Questions


Among the thousands of cryptocurrencies, tokens, NFTs, etc., Bitcoin is the only finite product in this cryptocurrency world. It has been continuously used for countless transactions since January 2009. At the same time, it is the currency with the largest capitalization and thus the most stable behavior in the cryptocurrency market. If you seriously invest in anything in crypto, in our opinion, it is Bitcoin.

Binance is the largest exchange. As such, we believe it is also the most stable and secure. At the same time, it offers some of the best if not the best transaction commissions.

Using this link to set up a new account on Binance, we will receive, both you and us, $100 USDT each for transaction fees.

Well worth it!

A detailed description of the programs can be found here.

Binance specifies that the minimum amount of funds in the futures account for operations to be executed is 0.001 BTC or USDT equivalent.

We recommend a minimum amount of 0.002 BTC (or USDT equivalent) due to significant changes in the price of BTC.

Lack of enough funds will result in no transactions - our bot will issue orders, but they will not be executed by the Binance exchange.

A huuuge one. Emotions have been a so far unexplored potential in behavioral finance. Until the research conducted among our scientists - linguists, psychologists and neuroscientists. We are probably the only company in the world that can measure emotions with as much accuracy as the average statistical human being.

For those who want to explore the topic more deeply, we discuss this in great detail in the following presentation - you are welcome.

The Exchange allows you to give appropriate permissions to the account key user. In the instructions you will receive from us, we show you which options you should check so that the bot can perform only the necessary operations on your account.

For example, the bot cannot withdraw funds, transfer to another account, use currencies other than those indicated, perform Spot trading operations etc.

Technical issues

It's easy.

It's easy.

  • We connect to your account through API keys.
  • You receive from us a preview of the bot’s trades with results and its history.
  • You get full support from us, both at the start and during the use of the bot.

No. We have not activated the lever at this point. The profits made by the bot seem very solid to us and we do not want to increase the risk of the operation at this point. Safety above all.

We put quality, over quantity. We worked on the strategy for Bitcoin for more than 2 years. It's a huge process, starting from collecting mentions through emotion analysis, learning a machine learning model, to developing the right signals for the bot.

We already offer predictive signals for 15 more cryptocurrencies at this point. For now, they are available via API.


In a nutshell, the scheme is as follows:

  1. The bot reads all mentions made in a predetermined few hours for BTC. Sometimes this is as high as 20,000 mentions per hour.
  2. The bot checks a certain amount of recent stock market data.
  3. From the mentions are emotions read.
  4. A machine learning model, trained on historical data for several years, predicts the average price for the next hour based on alternative data (emotion from mentions) and financial data (stock market data).
  5. The predicted price is converted into a specific buy or sell signal, taking into account the position currently taken by the bot (stable / crypto). A number of parameters must be met for the bot to perform the operation.


In the case of longs, the situation is obvious - the bot, assuming the growth of BTC, buys it in order to sell it the moment it receives an advance signal of trend reversal.

If, on the other hand, the bot assumes that the price of BTC will fall and at the same time holds BTC, it will sell it, buying it back at the next opportunity at a lower price. In this way, the bot makes a typical short transaction, without the participation of the exchange as a paid intermediary for this type of operation. The effect is the same, but with the additional commission saved on the short.

Currently, the bot decides every hour about the next operation, based on the predicted future price of BTC and the historical OHLC values (open, high, low, and close).


  1. not every prediction is hit. We have a high efficiency of the average closing price, but there are still missed predictions. It is not always emotion that decides the price, there are times when the market reacts to specific events rather than general sentiment.
  2. moreover, we predict the AVERAGE closing price. There is still a long way from this value to the issued signal on the market (with a specific value and at a specific time). There, too, something can go wrong.
  3. not always the operation issued by the bot (BUY_ORDER or SELL_ORDER) will be realized on the market (there will be a buyer or seller on the other side of the market),
  4. not always the increase or decrease in value is in accordance with the bot's minimum assumptions.


it is a good idea not to look at the bot every now and then and wonder why it made such a decision and not another. After all, I would have done it much better! In most cases, you wouldn't. Backward analysis is always effective;), but in a particular situation you know even less than the bot because:
you haven't analyzed thousands of tweets in the last hour for market sentiment,

  1. you don't have a clear strategy for when to enter and when to exit a position,
  2. you are driven by your personal emotion, it's your money after all!
  3. emotions in the market are a bad advisor.

Run the bot and take a look once in a while. Even if the bot lost, see if it lost more than the BTC itself. And if it gained, enjoy it. And let it do its job 🙂

First, in what they are similar:

  1. Both strategies analyze the same mentions and the same BTC rates, in the same way
  2. Both strategies make hourly decisions about the next step

There are a few key differences between the two that you should know about:

  1. A conservative strategy is designed for a bear market or sideways trend. It enters positions much less frequently (sometimes several times a month)
  2. Dynamic strategy can enter a position sometimes more than a dozen times in an hour
  3. Entry and exit criteria are set differently. In the conservative approach are less "risky", because the market is less predictable (frequent changes in the direction of the trend)