A Search Chatbot. Smart Search. A Protocol Bot

Practically, every day we need to search for something on the Internet, to get information about this or that, to find something we are interested in, to buy services or goods, spending or rather wasting sometimes a lot of time.

For example you need to find:

  1. A free or fee-based API workable in the JSON format to select synonyms for your chatbot;
  2. A vacation retreat to rent for three persons for a weekend at the price range of 10000 rub, and not further from Moscow than 50 km, so that it would not take you longer than 2 hours to get there.
  3. An auto service not far from your home, open on weekends, where they could do alignment for Lexus GS 300.
  4. A diode laser engraver for the price within 15 000 rub to be delivered to the office next day.

I Google searched for that information, then made orders and bookings by writing or giving phone calls, altogether spending on it about 10 hours of my time.

I am not fond of giving phone calls, and sometimes it takes too long to get answers to your email requests.

A chatbot could become a good assistant in that matter. But a universal protocol is needed for the chatbot to be able to request information for the user.

Let’s consider another example in the JSON format.

Protocol description:


Bot_name: #endurance,

General Variables: #replybefore, #phone, #email

Request Variables: #guests, #duration, #arrival, #departure, #pricelimit, #location, #range, #timetodrive, #link, #misc


An example of the request format:


Bot_name: #endurance,

General Variables: #replybefore =02/02/2017, #phone=+79162254302, #email=gf@EnduranceRobots.com

Request Variables: #guests=3, #durationday=2, #arrival=03/03/2017, #departure=03/05/2017, #type=resort, #pricelimit=10000RUR, #location=Moscowarea, #range=50km, #timetodrive=2hours


An example of the response format:


Bot_name: #hotelgreatnothern,

#replybefore =02/02/2017, #phone=+79162254302, #email=gf@EnduranceRobots.com

#guests=3, #durationday=2, #arrival=03/03/2017, #departure=03/05/2017, #type=resort, #pricelimit=85000RUR, #location=Moscowarea, #range=30km, #timetodrive=1,5hours, #stars = 4, #misc=suite, #more = free wifi, free parking, free swimming pool and sauna, #website = www.booking.com/pictures,


Thus, a chatbot could find only those variants, which fully corresponded to the chosen criteria, and offer the most optimal solution to the user.

A Request Interface:

A user writes a request in a free form, then a request is divided into variables with the given criteria. The more specific is the request the more exact and relevant is the bot’s response.

The request recognition is dynamic.

This will allow effectively get and analyze the information. In other words, a user will no longer be limited to those resources, and filters that offer products or services search. That will allow users to define their search criteria and receive responses on the basis of the given search criteria.

Advantages as compared to the present forms and filters:

  1. A dynamic response about the presence of absence of the services or goods corresponding to the given criteria.
  2. A possibility to create one’s own criteria for selection and receipt of this or that response.
  3. A possibility to get the best offer on the market.

How it works now:

  1. A chatbot is created to automatically send out requests via Google to the given number of resources according to the key request. In other words, Google results are not so important as the bot can request relevant information on 10, 20,100 different sites.
  2. The chatbot sends out a standard request to all the resources via: email or texting.
  3. Gets responses, arranges them and presents to the owner.

How it will work soon:

  1. A new chatbot will send requests to any number of resources.
  2. A chatbot can send a request not only to a resource but immediately to the chatbot of the resource. In a word, chatbots will communicate with chatbots to get the relevant information and present it to the user.

Thus, thanks to the chatbots, the user will have a full access to the information he needs.

In some cases such aggregators as Skyscanner, Booking, Expedia are good enough to present structured information with the given criteria to the user. But the absence of aggregators in other cases makes access to the necessary information much more difficult.

We believe that this kind of search bots will allow in the future making any complicated requests and finding necessary information for the users. Chatbots will make their own independent selection of the information for users, thus protecting them from irrelevant data, and saving their time.

Endurance plans to present the first pilot bot in the first quarter of 2017. The chatbot will collect all the emails on the basis of the given search criteria and will send out standard requests in order to get responses.

More information on EnduranceRobots.com

Or by email: gf@EnduranceRobots.com