Company: GOVWISE is an AI-powered procure-to-pay platform for the public sector, presented as a web application featuring a variety of dashboards that support analytics over public procurement data. Underneath this presentation layer, and supported on a robust infrastructure and clear data/software management principles, the GOVWISE product has multiple components that integrate developments relating to areas like Machine Learning (ML), Natural Language Processing (NLP), or Information Retrieval (IR).
Automated recommendation algorithm for Leads
Develop an automated recommendation algorithm of Leads (New Tenders and Predictable Contracts) based on i) Past Contracts of the Company (GovWise´s user) and ii) User inputs accordingly to the usage in GovWise´s platform (real time model training). For context, what GovWise wants to have is like the "youtube´s suggestion algorithm". Some of the data points available are:
- Historic of Tenders and Contracts with tender description, classifications (GPV, CPV), price, buyer, supplier;
- Tenders/Contracts Documents (eg: PDF).
Clustering tenders and companies
Develop a clustering algorithm that automatically separates i) Companies ii) Tenders/Contracts into different clusters, based on how "similar" they are.For example, one cluster will be IT, another will be Health, another construction, each with respective sub-clusters (Inside IT we can imagine software companies will be separated from hardware, etc). The goal is that there is no input from the user on how many clusters there are, or what their names are, the algorithm should figure this out. To accomplish the above, a large structured database will be provided, consisting of information on Tenders and Contracts, with tender description, classifications (GPV, CPV), price, buyer, and supplier.
Contact: Francisco Figueiredo ( francisco@govwise.eu )
Assessing similarity between tenders and contracts
Creation of an algorithm to calculate similarities between Tenders and between Contracts, with an emphasis on optimizing the performance of the same, taking into account the current volume of data and the expected volume of data.
Contact: Francisco Figueiredo ( francisco@govwise.eu )