Contract Fit Project


Challenge relevance

Transparency of international organizations budgets

Project idea (elevator pitch)

To provide companies actionable information on contract opportunities with multilateral organisations

The story behind the project

A lot of companies lack information about where and how much procurement opportunities are available
following questions we want to answer for suppliers:

  • to which multilateral organization should I apply?
  • how many funds will I probably get?
  • what where the procurement opportunities in the past for companies like mine?

Available datasets used

We used all the available procurement data of UN, World Bank and NATO available on Because of time and resource constrains, we focused on the UN

Some additional resources from Sicco: The UN already has some information for companies to obtain contracts you can use this as inspiration on what is already done, what is missing and use it to for NATO and World Bank as well.


Our goal was to add information about the companies that received the funds by adding size, profit, revenues. however, this information was very hard to get because there is no exhaustive portal available to find the information on an international level. We decided to only look at the contracters in France and use SIRENE to match suppliers with additional data, since we had access to this portal and since France received large funds from the bilateral organizations.

initial process:

  1. Assess types of procedures ( political analysis & legal)
    • Criteria
    • Timeline
    • success rate / number of bidders
    • objective criteria/ attribution criteria
  2. Analyse openmultilaterals data, and add companies information (data analysis, match suppliers of openmultilaterals data with supplier data of sirene)
  3. Find discrepancies between the two, and create the relevant graphs ( data analysis)
    • have an overview of where the most discrepancies are
  4. Design a website for a company to take strategic decisions ( UX/UI design, web development)

Are you fit for this IO ?
How likely to have contract with IO
Target country : Borrower country?


  • Types of project?
  • Size of project ?
  • What organisation gave contract on this countries?


problem we faced: does not provide us with detailed information about suppliers (revenue, size, profit) so we cannot match anything from it with the suppliers available in the openmultilaterals datasets
->we agreed on using the dataset availabe only and look at how we can use the information in the data to communicate how much contracts were given per country and sector in each area so that contracters will have an idea where to apply to. since we cannot match the datasets of UN, WB, NATO, for the moment, we will only look at the UN data since it is the most exhaustive. this should give an idea about the concept we have and can be developed further in the future.

adapted process:

  1. finding available information about the suppliers that we can use from the dataset in openmultilaterals
  2. finding available information of the suppliers via national registers (sirene), opendata…
  3. research about procurement procedures of UN
  4. searching for opportunities to communicate the informational understandably on a homepage
  5. designing and creating the homepage

Additionnal notes :


List of UN divisions available in the dataset:

un_organisat |
ion | Freq. Percent Cum.
ECA | 142 0.06 0.06
ECLAC | 322 0.13 0.19
ESCAP | 358 0.15 0.34
ESCWA | 147 0.06 0.40
FAO | 10,329 4.30 4.71
IAEA | 3,919 1.63 6.34
IFAD | 1,499 0.62 6.96
ILO | 4,111 1.71 8.68
IMF | 509 0.21 8.89
IMO | 303 0.13 9.02
IOM | 4,594 1.91 10.93
ITC | 472 0.20 11.13
ITU | 519 0.22 11.34
OPCW | 491 0.20 11.55
PAHO | 4,142 1.73 13.27
UN-ICTY/MICT | 108 0.04 13.32
UNAIDS | 589 0.25 13.56
UNAKRT | 14 0.01 13.57
UNDP | 52,535 21.89 35.45
UNECA | 457 0.19 35.65
UNESCO | 5,255 2.19 37.83
UNFCCC | 288 0.12 37.95
UNFPA | 8,389 3.49 41.45
UNHCR | 17,474 7.28 48.73
UNICEF | 28,605 11.92 60.65
UNIDO | 3,023 1.26 61.91
UNOG | 2,176 0.91 62.81
UNON | 1,478 0.62 63.43
UNOPS | 19,878 8.28 71.71
UNOV | 1,104 0.46 72.17
UNPD | 26,457 11.02 83.19
UNRWA | 5,562 2.32 85.51
UNU | 203 0.08 85.59
UNV | 262 0.11 85.70
UNWOMEN | 2,007 0.84 86.54
UNWTO | 88 0.04 86.58
UPU | 284 0.12 86.69
WFP | 25,879 10.78 97.47
WHO | 4,863 2.03 99.50
WIPO | 718 0.30 99.80
WMO | 422 0.18 99.98
WTO | 58 0.02 100.00
Total | 240,033 100.00

scripf for descriping the agencies:

let agencies = {
“IMF” : {
“name”: “The International Monetary Fund”,
“description”: “IMF facilitates international monetary cooperation and financial stability, and provides a permanent forum for consultation, advice, and assistance on financial issues”,
“ILO” : {
“name”: “The International Labor Organization”,
“description”: “ILO formulates policies and programs to improve working conditions and employment opportunities, and defines international labor standards as guidelines for governments.”,
“logo”: “ILO.png”
“” : {
“name”: “”,
“description”: “”,



Company form questions :

Country of origin
(might be compared at the company group level: emerging, OECD, memberstate of this organisation)

Size of the company
Sales, number of employees

Field of action

Information you get back

Are you fit for this OI ?
How likely to have contract with OI
Target country : Borrower country?

Types of project?
Size of project ?
What organisation gave contract on this countries?
Other companies
tell why service within the UN
organisation within World bank

How many people apply to bid ?
Success rate is critical

Information we need to find

  • success rate ?
  • list of agencies, headquarters, logos
  • types of procedures, lenghth
  • redact the description (forum )

To the interface :

  • why we recommend it
  • country selector
  • heatmap


Slides and Pitch Preparation:


Please share the link to everyone, I cannot open it.
Make a public link so that I can view from the laptop connected to the screen


feedback on the presentation:

  • do not point on the slides bc we will not have screens in final session
  • stick to 4min, if it is less, then add content
  • be more confident in the end when describing the homepage
  • put legen into the graphs!!
  • slide 2: make graph more attentious
  • slide 3: have the video more aligned to slide size

Feedback to presentation in person:
2nd slide: what did we do?
3rd: the impact

link the graph to the website
add the source of the data
broader impact ->bold statement
word “recommendation”: replace with “top 3 contract providers”


Key words:






Multilateral organizations


Match, fit

Title ideas:


Find your contract

Opportunities through transparency

Strategic contracting

Bingo/eureka Procurement

Description of our “tool”:

The tool that companies use to choose multilateral organizations for successful procurement contracting




Hello guys! Here is the link of our Python Script for the Contract Fit Project:

I also want to share with you the link of the additional UN database that we used for geographical classification in our script:


Hi there! Here are our main problems we have encountered during these three days !

We have worked on the openstate’s database, which regroup information about three world organisations: UN, NATO and The World Bank. The data set includes information about the contracts signed by these organisations, the contractors, their countries of origins, the amount of money delivered for the contracts and the description of the contract.

Problem faced:
Our Sample was composed of 240000 contracts on 7 years from 2010 to 2017.
First we observed that some contracts (around 1000) were kind of debt contract. (negative or null amount).
Many firms have missing informations about the amount and the description of the contracts.

For our website we wanted a multiple choice for sector. We wanted to classify the sector according to the UN classification but the data were too heterogeneous in the description (sentence, language, unknown type of contract, numbers, symbols). We tried to analyse the most frequent terms. The function counted the most accurate words used in the description. But there are a lot of missing data, and our classification doesn’t cover all the sector in the UN classification. Our « sector » columns is biased and incomplete, and it is a limit of our work. Then we associated the sector with the description it suited the most. Our objective was to attribute to each contract a single and simple sector, so that we could to statistical analysis on it.

We wanted to highlight how « description » is a blurred category and do not goes in the sense of transparency. Many of the description were incomprehensible for the public and for us, so difficult to classify in sectors.


The team !


Code repository
Live demo ( medical data in Sub-Saharian countries )


Thank you for sharing the repository :sunny:


Thank you so much for the excellent documentation effort!