Smart Contracting Is A Data Challenge

Making government procurement smarter is a data challenge. It is a challenge to fight corruption, to test ideas for improving the current system, or just to have some fun with a vast array of practical datasets. Is is also why DSP is holding the Smart Contracting Data Challenge.

Challenge Goals

  1. Explore new ways to improve government procurement processes and quality
  2. Transform the government procurement marketplace into economy stimulator.

Challenge Areas

  1. Procurement Chatbots
    • To develop a chatbot to guide suppliers and buyers through the complicated layers of procurement regulations and procedures.
    • If succeeded, the chatbot will act as a customer self-service, to reduce repetitive workloads in the Public Construction Commission.
    • Datasets: more than 3000 well-organized FAQs and related laws/regulations
  2. Open Innovation
    • To help buyers find quality suppliers, and vice versa.
    • Explore innovation ideas in government procurement
    • Dataset: Historical award records, compiled in Open Contracting Data Standard (OCDS)-compatible schema.

Last month we held a pre-event seminar. Two speakers shared their experience in analyzing government procurement data, from the perspective of auditing, and from the perspective of business research. We also had a speaker telling his sweet-and-bitter story in developing chatbots.

This data challenge is unique in several ways:

  • Multi-stakeholder partnership
  • Domain expertise
    • We are lucky to have several domain experts acting as consultants to help contestants who are fluent in the languages of data and programming, but not government contracting.
  • Focus on sharing and learning, not competing
    • Although the event is called a challenge, we do not implement any rules to judge performance of participating teams.
    • Instead, we offer co-working sessions and a group chatting solution (Slack) to facilitate communications and collaboration.

The Smart Contracting Data Challenge is now in progress, and we are expecting results in January.