UA Anti Spam Bot Case Study

Chatbot that fights fraudulent messages and strategic information leaks

Effective communication has been challenging since the onset of the Ukraine war. MOCG embarked on a project initiated by our Master, Dmytro Vakulenko, to develop a chatbot that automatically filters out and deletes dangerous content, strategic data, and sensitive locations. We aimed to ensure only appropriate information is shared to protect and control the news flow during these difficult times.

Try Now

1 400 000 of users that moderate their communities with Anti Spam Bot
17 000 active chats for now, including channels and private messages
72% of chats use blockage of fraudulent messages

Dmytro Vakulenko Conversational Frontend Engineer in Master Of Code Global

We take pride in that our product, UA Anti Spam Bot, has been recommended by the Ministry of Digital Transformation of Ukraine as a reliable tool for checking and moderating Telegram groups. Furthermore, we have earned the trust of 1,400,000 users who have chosen our bot to effectively moderate their communities.

Where It All Started

The idea for the bot was born after 24.02.2022, when Russian invasion of Ukraine had happened. Communication channels were all over the place; often, they contained false information but also harmful, and deceptive information. People didn’t know what data could be shared, resulting in a really dangerous situation for Ukrainians. Sometimes admins of news channels on social media couldn’t manage all this information in time. That’s why automation seemed like the ultimate option.

Dmytro Vakulenko, Conversational Frontend Engineer at Master Of Code Global, came up with an idea of a bot that would make our informational space a safe place for everyone. To identify fraudulent messages, he gathered the most suspicious keywords and locations: when there are two or more of them, the message is immediately blocked.

Challenge

What were the main difficulties in developing our Anti Spam Bot?

As the bot was created from scratch, our team had to manually gather strategic information and high-risk words. However, we had volunteers that helped us with this mission, and they were not only technical experts but also ordinary people

When the bot was used by a large audience, it couldn’t work properly. This is why we decided to opt for a separate server that helped answer the questions and receive messages, making the process easier.

The last challenge we solved was the lack of context for chatbot. Sometimes the information was harmless, but the bot deleted it because of certain keywords. So, we developed a neural network that could distinguish safe messages from sensitive ones. To implement it our specialists collected 4000 examples of possible strategic data.

What We Created

What We Created

Master Of Code Global developed a bot that blocks high-risk words and strategic information. The chatbot doesn’t let them share fraudulent messages which may also include card numbers and URLs. Explicit content, such as 18+ stickers, videos, and GIFs is also being deleted. An additional plus for Ukrainians is also the blockage of Russian language. The first time you use it, you either receive a warning or your page is automatically blocked. The chatbot also detects antisemitism and deletes all extremist messages.

What Is Our Anti Spam Bot Capable of?

Fight harmful comments that expose the location of the Armed Forces of Ukraine, blockposts, places of explosions, etc.
Protect from phishing and frauds, including fake charity and help from organizations.
Block the adult (18+) content, including gifs, videos and photos.
Automatically delete the greeting and goodbye messages to preserve the confidentiality of your chat.