Want to know how to build a conversational ‘chatbot’? Here are some tools

4 min reading
APIs / 11 May 2018
Want to know how to build a conversational ‘chatbot’? Here are some tools
Want to know how to build a conversational ‘chatbot’? Here are some tools

BBVA API Market

Four of the most widely used tools for creating simple conversational chatbots.

Chatbots are the current hot trend in the technology world. They may be becoming better known among the general public, but conversational bots are not a recent development. Some companies have been experimenting with the developments that today underpin everything that is being done with bots for a long time: basically, this involves processing natural language to enable a machine to have an intelligent conversation with the user. This would be totally impossible without these developments.

When we look at the natural evolution of IT and how users relate to it, it is easy to understand what some of the world’s biggest companies are up to. In the mid-2000s, smartphones entered our lives, with their mobile operating systems (iOS and Android), native applications and native mobile-user experiences, all on the client side.

Current trends are still focused on smartphones, but now on messaging apps, where direct contact through chatbots has become king. The diagram below shows how technology has developed since the 80s

This has been made possible by the enormous push to natural language processing, automatic learning, virtual assistants and artificial language in general by companies such as Facebook and Google.

A number of smaller companies have also understood the real business opportunities in the creation of platforms for easy creation of conversational chatbots, some of which do not need you to write a single line of code. We describe some of the most interesting platforms for creating bots below, based on the need for flexibility in introducing an MVP:

1. Api.ai

Api.ai must be doing something right to catch the eye of Google. The creators of this platform are fostering a space for the creation of bots for third-party scenarios, such as: Slack, the project management and internal communication project for companies; Facebook Messenger, the Menlo Park company’s chat program; Kik, an instant messaging app competing with Facebook Messenger; other players in the sector, such as WhatsApp, Telegram and WeChat; and more purely social products, such as Snapchat.

The best thing about API.ai is that you can create a chatbot for Slack or Facebook Messenger without any programming knowledge: you do not need to know a programming language or memorize any commands. The idea is to create the bot, generate the paths for different messages, implement integration in a messaging platform and publish. 

2. Motion.ai

Motion.ai is a simple chatbot creation platform for two of the leading messaging apps – Facebook Messenger and Slack – of the moment. It can also create bots for browsing, email and SMS. There is a free version, which can develop bots with up to 1,000 messages per month; and three paid versions: one for $15 per month (5 bots and 5,000 messages); one for $50 (with 25 bots and 20,000 messages per month) and the most expensive, at $100 per month (50 bots and 50,000 messages per month). Motion.ai also offers its own API REST for integration with its own interface.

All you have to do to create your bot is register, enter the site and go to the ‘My Bots’ menu on the left. Here you can create as many chatbots as you want, by clicking ‘Create Bot’ and selecting the channel where you want to create your product (for Facebook Messenger, Slack or a bot for email, a browser or SMS). All the technical information you need to start to create your chatbot is available on the documentation page. 

Once you have named your bot, you can go to the edit page and configure the modules for the chatbot’s user conversations. Each of these modules represents a specific reply by the bot, which can be edited using the ‘Configuration’ button. These modules form the various elements in the conversation pipelines, which are represented by connections shown by colored lines and a flow chart. Motion.ai offers many types of modules, and you can choose the type of content you want by clicking ‘Add Module’. It also has direct access to testing functionality for the bot and its metrics. 

3. Smooch.io

Smooch.io describes itself as an interface-creation platform for companies. It offers a wide range of integrations with third-party solutions, including all of the big boys (Facebook Messenger, LINE, Twillio, WeChat and Telegram), and both cell-phone operating systems, iOS and Android.

Smooch.io also offers its own API REST for personalized integration. The idea is to offer a seamless customer experience (incorporating all types of content formats – text, images, gifs, videos, etc. and visual elements, such as buttons), providing developers with the flexibility and scalability needed to connect chatbots to Customer Relationship Management (CRM) systems.

This is one of the things that really sets Smooch.io apart: it can connect conversational interfaces with a company’s CRM system or internal customer-management platform. It also offers a full range of metrics and reports on the combined working of these platforms. 

4. Gupshup.io

Gupshup.io is a chatbot development platform for practically all of the messaging tools, social networks and VoIP solutions on the market, Facebook, Twitter, Telegram, Slack, Hipchat, Twillio, Line and Skype, etc., and will be expanding to Google Hangouts, Viber and Kakao in future.

Gupshup.io offers a simple code-editor solution (this is a major difference compared to the other platforms – you have to program), a publication service and a testing system for each chatbot

Gupshup.io handles the installation of the libraries and packages needed to create the bots with no need for input from the developer, and a secure server for roll out, meaning you don’t need your own server.

It then leaves the creation of the functional logic for the chatbot in the hands of the programmers. Gupshup bots are developed in the upper part of the Amazon AWS Lambda service, in a serverless computing service that executes code in response to events. It also offers a testing service, enabling developers to test their product as if they were already live on Slack, Telegram or Facebook Messenger. 

It may interest you