hu:toma Bot Building 101 ( Just add .txt )
Building a “chatbot” is currently the on-trend thing to do, but why? Corporate innovation teams, SME’s and freelancers alike see the opportunity in adding conversational functionality to their web & mobile applications.
We’re fast becoming a culture that spends a lot of time in chat applications, and unsurprisingly, businesses want to adapt and take advantage of this. Mark Zuckerberg said in 2014 that “messaging is one of the few things that people do more than social networking.” According to SimilarWeb, for the average american, over 23 minutes a day are spent in chat applications.
Slack, the largest workplace chat application, was built with bots at its core. Artificially intelligent (AI) bots have the power to improve user experience, increase user productivity, and customer engagement by delivering contextual information.
This trend has given rise to numerous companies creating chatbots, digital assistants, and conversational user interfaces. These chatbots can do anything from scheduling meetings like x.ai to giving you the weather forecast like Poncho. We at hu:toma have created a platform for 3rd parties to build these domain specific chat bots and distribute them.
Getting started with hu:toma
Creating an AI is as simple as uploading a text training file. It’s that easy!
I’ve decided to create a Police Bot using some frequently asked questions from the Metropolitan Police’s website.
To set up your chatbot, add your training file to a Text Editor and save. Make sure the questions and answer span only one line each.
Then upload the training file and let hu:toma’s deep learning technology based on Recurring Neural Networks (RNN) do all the work. There is no need to manually create conversation trees, which can quickly become incredibly complex for more than a limited conversation.
You can choose how often the AI can create new answers that fall outside of your data set / training file, and can also choose whether the AI will learn from its users or not.
Once Phase 1 of the training is complete, you will be able to chat with your AI in the platform chat window. As you chat, you can also look at the JSON file that is created if you wish to test. Phase 2 of the training is the “deep learning in action.” This allows the AI to create new answers based on the training file. We’re busy building integrations for all the major chat platforms, so you can be assured that you’ll be able to communicate with all of your end users.
AND hu:toma sets itself further apart by allowing users to publish to the Bot Store, a marketplace that allows developers to monetise their creations, increase their proliferation and drive sales through integrations with 3rd party applications.
The hu:toma Bot Store allows you to combine multiple AI’s that are trained with data from domain experts to create your ideal user experience. Imagine allowing your users to order coffee from Starbucks Coffee, get it delivered by Stuart, receive their favourite content daily from Medium, and book a spot on their Spinning Class at Pure Gym, all from within a single hu:toma powered chatbot.
Be first to use the product by joining our waitlist.
