How to create an army of Bots

How To Create An Army of Bots

A Developer’s Perspective

Imagine a world, where a centralized system processes a (text) input from the user- as a chat message- and acts based on these commands. Retrieving bank balance; recalling flight tickets; filing a complaint against a service; entering them into a draw for a competition etc. The possibilities are literally limitless!

One month into 2016 and the biggest platforms in the world are releasing- or have released- APIs to enable 3rd party developers to develop bots and responders; a pretty clear indication of what to look forward to for the months to come!

While some platforms already provided such developer access, the fact that giants like Facebook are also exposing their messaging capabilities, goes to show how important it is to consume those APIs and develop tools that would enrich the user experience for your brands products, services and marketing campaigns.

Of all the APIs being open to business, perhaps the most anticipated are Whatsapp APIs. We don’t really know what are they going to release, how, and when, but looking at how Telegram have opened up their APIs we can get some ideas. An auto responding bot is begging to be created using these APIs!

Enter Brndstr Bots.

The main challenges here would be:

  • To create the system; a system where a knowledge base could reside, and trigger actions based on the user’s interaction with these different bots. Such a system would be independent of the platform being used for communication. To do so, we would have to write a specific communication layer for each platform, enabling communication back and forth from the backend system to the chatting platform. Building such layers, would be a lengthy but fairly straightforward procedure i.e. to consume and conform to APIs for each platform.
  • To build a knowledge base and wrap it around a logical layer to perform user actions; essentially, an AI challenge. This process would come with a host of challenges to solve and optimize, such as committing information to the knowledge base, querying and deriving the desired results.

A bot is nothing but a clever engineering team behind it. An army though, would require a whole lot more than just clever engineering. The right architecture for scaling up, coupled with just the right backend functionality to maintain and query a knowledge base and to help make decisions, are critical.

At Brndstr, we do have a few solutions and best case practices to these challenges, since we have been working with Twitter APIs, but generalizing a targeted system will definitely need a more clever design, and we’re excited to work on that in the coming few months! We have the collective knowledge and experience, so it’s just a matter of time until our army comes to fruition!

— 
#Jasim / The iOS Dude

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.