As the makers of an open-source chatbot framework, we are always excited and inspired by what our users are building. Today, we are interviewing Yacine, who created SimSimi Algerian, a very popular chatbot with over 80.000 active users, and more than 3 million messages every month.
Hello Yacine, it's nice to have you today on the CSML blog! Can you tell us a bit about yourself and your background?
Hi, my name is Yacine and I am a chatbot builder from Algeria. I build and experiment with chatbots as a hobby and to be able to know the limits they can reach. I learned a lot over the past years and I have good experience in dealing with users and how they interact with chatbots through the Messenger platform.
What is SimSimi, your most popular chatbot?
Simsimi Algerian was a Messenger chatbot first introduced in 2019. It is a conversational chatbot that can interact with the user in a human-like manner and provides a great experience to show the capabilities that the chatbot can achieve. Since July 2021, more than 80,000 users have joined to talk to Simsimi, and more than 4 million messages have been exchanged during that period.
How did you get the idea to create SimSimi?
I am inspired by a lot of chatbots like ELIZA and other similar projects, so I decided to build a chatbot that mimics the way people speak and act on social media. At the same time, SimSimi acts as a friend of the user, who can talk to him about many topics.
Since 2019, there have been several versions, but there were things that the chatbot could not do, such as remembering things and using conditions to respond to the user.
One day, while I was looking for chatbots building websites, I read about CSML on Reddit. I signed up and started reading the documentation. It was very easy because it didn't require me to have much programming experience, so I connected CSML with Messenger, I wrote my first code and got the first response back from CSML at that time.
Then I started redeveloping SimSimi while learning CSML at the same time. It only took me one week to write SimSimi responses and learn the basics, which is amazing! Then I started experimenting and discovering the things I could achieve with CSML. I really liked all that I could do, like remembering names, places and sending cards or pictures easily. That's not making the chatbot smarter, but it does make it more efficient in handling user messages!
Do you use any NLP or external service/API?
In general, I focused on using CSML because it supports the use of external NLPs service. That's why I use Dialogflow as an NLP provider, which helped me a lot in the success of the chatbot. I didn't try too many NLP providers, but I used Dialogflow for its ease of use and navigating between the phrases and intents
However in my opinion, using an NLP service is not always necessary, because CSML already supports it and can do most of the tasks.
How do you explain SimSimi's success? What makes it such a popular chatbot?
SimSimi's success was for many reasons. In general, I have always worked hard to make SimSimi conversations and the messages that he sends unexpected and funny in most cases, so I rely on the users taking a screenshot of the conversation and sharing it with their friends. Most of the time they ask him what he thinks of someone or does he know this place, then he will respond with a good or nice answer that makes the user happy and has a desire to share the response with his friends, and this is the strategy that I adopted made SimSimi famous on social media and successful.
It took me a little patience, but it was worth it, and I think that the home quarantine that took place in early 2020 made many people use social media more and got to know SimSimi at that time.
Note: It took me a whole year to reach 5k likes at first and only one month to reach 30k!
What are your plans for the future with SimSimi? Do you have other projects?
Since SimSimi has achieved such success in the short period of time I started using CSML. I plan to improve it further, add some responses that are missing and then move on to work on another project that supports many languages and provides an exciting conversational experience, which I have discovered that I can achieve with a little work in CSML!
In my opinion, CSML is the best engine I used to make the most complex chatbots because it provided me with a language to program the way the chatbot behaves easily and quickly. I am very grateful to the team who helped me with most things on the Slack platform and to the developers who shared their experiences, problems and solutions, and their experience with CSML.
That's very kind of you Yacine! We are glad we have been able to help you building this fun project! Thanks a lot for sharing your experience with the community. We do wish you all the best in your chatbot projects!