DATA AGGREGATION ROBOTS
As a part of the business development strategy we will not wait for our eventual clients till they come to visit our portal, think and register to our services. We will upload the data about tenders by ourselves. Same we will do with solutions and products of smart technology suppliers. We will scan, analyse, rate and upload product data to our portal by ourselves. We plan to make it by sets of data aggregation robots that will scan public web portals and transfer data to our platform automatically. For tender information it will be fully automated process, for smart solutions and products it will be semi-automated process with interaction of our employees who review the data, create custom graphics, execute initial rating, etc.
Aggregation software robots scan portals and store data. During processing assign to tenders categories and tags.
There are dozens of different tendering web portals over the Internet serving some data and providing some services. Each portal have it's own front-end with an unique structure and layout, tenders data are represented in different ways, include different sets of attributes and documents, filtered in different ways. Databases running in the backend of these portals are also designed differently. Still there are major similarities because regardless of the way how tendering data is represented, it is basically describes the same or very similar entities.
Simpler, if we take 20 random independent tendering platforms and assume that each tender contains about 120 fields to fill plus a few attachments related to the project, and then compare these data, we'll find out that around 75% of attributes can be matched perfectly, the rest is either not important because other platforms don’t need it at all, or vise versa contain some sort of specific yet important information that worth being included as a mandatory field in our Master list of attributes. Our software robots scan, collect, sort and store this data content in our Tender platform. Each robot will have specialised configuration for exacting the data from the particular tender platform and collect it from chaotic form into our database. During processing, tags and categories will be assigned to tender data – either automatically or by data stewards.
There are two Data Input (DI) parts in the data aggregation architecture of the solution(read scheme from bottom). These are two databases where source tender and smart solutions data arrives initially.
DI 1 collects information from two main input channels. The first one is a manual data load by Tender Creator client by registering a new tender. The second is automated data load from other public tender portals by our software robots. All data together give us the "Buyer" content part.
DI 2 collects data from three independent input channels. All channels work either semi-automatically or manually due to extreme content diversity. The first one is input by a Smart products suppliers directly by their registration of company and products on tender portal. They might do that either to apply for a tender or just actingproactively in order to expose and promote their products or technologies. The second one input channel are Agencies Data. Catalogues of companies, moderated by an agencies. We will buy data from them, adjust and load into product database. The third input channel are our own employees who will research solutions on webs, consult with suppliers and load data into our catalogue. All data together give us the "Seller" content part. All these data are further processed in the CAT database and business logic that properly tag them and categorise.
Collected and processed data are further used as a main content for three Outputs. Output - Tender products - 1 cover initial data for creation of our tender mediation products. It will be used for digital marketing first and for selling a consultation services providing by a Smart City Enterprise company. Output 2 is the main content of our Tendering Platform. A data the will be available for clientsin public or paid catalogues. Output 3 is a second part of our business plans to create Business Intelligence reporting system providing commercial reports for business, forecasting and financial needs.