By now, it has become clear that bots will cause significant systemic changes in customer service, e-commerce and, in truth, in all aspects of software-human interaction. The current state of bots - bot-to-consumer, it means that bots communicate with people. But in the near future they will start talking not with us, but with other bots. A new era of bot-to-bot begins.

Imagine that the bot - for convenience, we will call her Annie - you need to answer the client's question. But information from its own server systems is not enough. Then Annie uses her artificial intelligence and spontaneously decides to turn to another bot in order to find the right answers. She receives information and transmits it to the client.
')
Death API?
Today, when two software systems can talk to each other, software developers need to integrate using APIs (application programming interfaces). This process is time consuming. That is why in the last few years services such as Zapier, Scribe and FTT have become popular. They provide exceptional interfaces for hundreds of applications, allowing you to connect, for example, your CRM system to distribution tools or analytics platform.
However, in the era of bot-to-bot, software applications can talk to each other’s systems, regardless of whether they have existing API integration. Of course, bot-to-bot communication will not use the exchange of large amounts of data, but it will create a special connection between, for example, user banking software and the online store. Banking software can talk to an online store bot and ask for a lost account: “My client needs an account to order 45678, can you provide it?”.
Big Final: bot-to-bot-to-consume
The beauty of bot-to-bot communication is that it will occur in plain English; it will be a conversation that everyone can understand. Provided that all conversations between my bot Annie and other bots are archived, I can go back and see how they came to a definite conclusion. In my banking example, when the account is lost, I can click on the "more" button and open Annie's conversation with the online store.
But further it is still more interesting. If my bot comes to a standstill in a conversation with another bot, she can call for help: “Hi, this is Annie, your financial bot. I talk to the supplier, but I have some difficulties in understanding what he says to me. ” I can join - of course, if I have a few free hours (as the bots' patience is unlimited) and rephrase Annie's question to get an answer from another bot. Further, Annie can continue the conversation and manage my affairs.
Semantic Web
Didn't we talk 10 years ago about joining one online service to another? What was it called? That's right, semantic web. Each website was going to be annotated using standard data formats, allowing other services to scan this data and use it in their business logic. I believe that bots will perform similar in the next 3-5 years and this means that all data will be formatted evenly. Instead, bots will demonstrate online services and data in plain English, allowing people and other bots to interact, even if they have never communicated before.
Convening all software developers
So, software developers, when you develop your platform for e-commerce, online marketing, finance, ERP (Enterprise Resource Planning) system or any other software solution, please consider implementing a smart bot other than your traditional API interface.
Link to the original article