As machine learning algorithms continue to grow, developers are constantly looking for ways to put AI into the consumer’s hands.
Mobile apps and websites have served as the conventional examples of front end development, but developers find it difficult to integrate AI into their platforms.
At GoFind, we believe this is where chatbots come in.
Chatbots are the only lightweight platform where every single layer of the application is powered through AI. So, we developed Anna the FashionBot, a chatbot that allows users browse across millions of products and shop through both text and visual AI search.
Implementing a Chatbot AI Stack
Starting from the front end, chatbots are able to leverage the power of NLP to make every input instinctive to the user. By doing so, chatbots are able to break to barriers of traditional user experience. Using the bot is more like talking to a friend rather than clicking through a series of screens.
Inputs of text or images flow naturally from chatbots into GoFind’s AI backend. With camera and gallery built into Messenger, images can easily be passed into our AI visual search. Text inputs are also processed to create the best text search query. The products returned by the backend are also simply displayed through carousels.
We also tune the results of every search based on each unique user. We are able to learn their preferences based on their previous inputs and interactions to deliver products they are interested in most.
By using machine learning at every part of the stack, we were able to make our chatbot fully powered by AI.
Perks of the Chatbot Platform
Besides AI, there are numerous other benefits of chatbots.
Being a lightweight platform, we were able to hack together prototypes of our chatbot in a couple days. The ease of Facebook Messenger development combined with the simplicity of handling simple POST requests significantly reduced our production time. Instead of building a mobile app from scratch, we were able to utilize the power of Messenger carousels and quick replies to build a clean and intuitive UI/UX.
Chatbots are also truly seamless across platforms, covering both mobile and desktop, spanning from IOS to Android. With only one application for every device, development time is heavily reduced and reach is significantly increased.
Unlike mobile and desktop apps, users are not required to install chatbots. This proved to be extremely helpful when we were at an AI conference, as people were able to access our chatbot in a matter of seconds. By being only a Facebook search away, a chatbot was the easiest way to share our AI.
Being built into Facebook, chatbots are also able to reach Facebook’s incredible global audience. By being built into Messenger, Facebook Ads and marketing are also extremely easy to integrate with chatbots. Messenger chatbots can also be easily built into other messaging platforms, as most of the input processing remains the same. We are currently working on expansion of FashionBot into Twitter and Kik.
Conclusion
Overall, the ease of production and integration of AI make chatbots the best platform for any AI engine.
At GoFind, we had an amazing time building our chatbot and we hope you will integrate chatbots into your next AI product.
Check us out at http://gofind.ai/
Check out FashionBot at http://m.me/annafashionbot
Thanks for reading!
