Behoover is mostly built on a standard LAMP service stack .
The AI system, however, stands out a bit from the rest of the system. It is highly specialized for its tasks (see below) and very fast. It was developed in house and coded in c++.
We automatically harvest company profiles
Python 3.x scripts crawl company websites using Selenium and turn them into large numbers of company profiles on Behoover. Text, email addresses, geography data and social media links are examples of data that is collected.
Other scripts crawl and add data from trade organizations, certification authorities and rating sites. The data collected is used to improve the company profiles even further.
We automatically categorize company profiles
The company profiles are automatically categorized using the AI system.
We help the customer write a professional request
Templates for common requests can easily be built using a library built on top of the twig templating system. Data is fed into those templates via our Javascript questionnaire system.
We automatically send requests to the best companies
The AI system will evaluate a customer request and start sending it to fitting companies (based on geography, categorization, ratings etc). It keeps sending until enough (usually three) companies have replied, or it runs out of good companies. If the latter happens it calls for (human) help.