At the digital conference The Future of AI & Big Data Analytics Rositsa Zaimova, Co-Founder & Partner at Dalberg Data Insights, give insights as on how to build inclusive data ecosystems for the global south.
Rositsa is Partner and Co-founder at Dalberg Data Insights (DDI) based in Kampala, Uganda. In the past 6 years, Rositsa has managed the design and development of data ecosystems and data-driven products in Sub-Saharan Africa (SSA) with focus on public health, gender, youth employment, agriculture, poverty. On day-to-day basis, she manages a complex ecosystem of data providers, data users, regulators and funders and has been advising digital and data strategies with governments and development partners that set the foundations for the Digital Transformation in SSA.
Under her leadership, DDI has been supporting the Ministry of Health in Uganda, in partnership with UNICEF and Rockefeller Foundation, to better understand the impact of the COVID19 pandemic on the continuity of essential health services (CEHS) at national and sub-national levels using public health data. As part of the project, the team has set up a governance structure and a technical infrastructure, which enabled the development of a CEHS app that fosters a data-driven culture in maintenance of CEHS within the Division of Health Information an MoH and its partners.
She has also led the launch of AIDA, the Action Insights Data platform in Uganda, which was developed in partnership with the United Nations Capital Development Fund (UNCDF) and the Belgian Government. The goal of the Platform is, first, to accelerate the use of private and public data sources for better decision making and, second, to create an entire data ecosystem for the country, while complying to regulations and ensuring privacy.
Rositsa was recently named one of Forbes’ 30 Under 30 Social entrepreneurs in Europe.
Building Inclusive Data Ecosystems for the Global South
1. Data Insights
Building Inclusive
Data Ecosystems for
the Global South
T H E F U T U R E O F A I A N D B I G D ATA A N A LY T I C S
R O S I T S A Z A I M O V A , C O - F O U N D E R A N D PA R T N E R
M AY 2 0 2 1
2. 1
01 Our purpose and approach at
Dalberg Data Insights
How we use AI for good
02
3. 2
Dalberg
Design
Dalberg
Catalyst
Dalberg
Implement
Dalberg
Data Insights
Dalberg
Advisors
Dalberg
Research
Our core businesses
Dalberg Advisors has strong strategic and rigorous analytical capabilities with deep knowledge and networks
across emerging markets. It works collaboratively across the public, private and philanthropic sectors to fuel
inclusive growth.
Dalberg Catalyst that brings to life multi-stakeholder initiatives to address today’s most critical social and
environmental challenges. Their goal is to take on urgent problems that have massive implications for our
collective future, and to anticipate and get ahead of emerging problems that are taking shape.
Dalberg Data Insights acts as a bridge between private data and
important social problems. It creates tools to interpret data and uncover
solutions in international development, humanitarian action, and social
impact.
Dalberg Research provides research and analysis to find actionable
solutions. It helps to answer questions about markets, consumers and
lifestyles across more than 20 African countries via quantitative and
qualitative studies. It specializes in the inventory, collection, and analysis of
primary research data.
Dalberg Design fosters creative solutions to enhance economic opportunity
and human potential, combining the best skills in human-centered design, rapid
prototyping and systems thinking to accelerate positive change in underserved
communities.
Dalberg Implement takes strategy recommendations and works alongside our clients to bring them to life. We
provide implementation services that marry strategy, innovative resource mobilization, and the deployment of
top talent to launch, scale-up, or transform impactful initiatives.
4. SAN FRANCISCO
JOHANNESBURG
NAIROBI
DAR ES SALAAM
DAKAR
LONDON COPENHAGEN
GENEVA
MUMBAI
NEW YORK
WASHINGTON DC
SINGAPORE
LAGOS
ABU DHABI
NEW DELHI
HONG KONG
PARIS
ADDIS ABABA
BRUSSELS
MEXICO KIGALI
ABIDJAN
SEATTLE
KAMPALA
3
24
Locations
90+
Countries we’ve worked in
58+
Nationalities of staff
80+
Languages spoken
2700+
Engagements globally
With a Global Presence…
5. We build inclusive data ecosystems across geographies and
sectors through governance, technology and community
4
PILLARS OF
INCLUSIVE
DATA
ECOSYTEM
Data governance
We develop the governance, protocol, and
technical frameworks to build data-driven
digital solutions when collaborating with a
variety of key stakeholders.
Community
We orchestrate inclusive data ecosystems
and communities of key stakeholders when
a new data-driven digital solution needs to
be sustainably institutionalized.
Technical tools
We leverage open-source technologies in
combination of private and public data sources to
create social and economic value by analyzing and
visualizing data insights that can be linked to a
clear set of actions.
The data governance is
critical in the data
ecosystems we create as it
ensures establishment and
adoption of the standards,
structures, and processes
required to successfully
design, develop, and
establish a new data-driven
digital solution.
The community activities ensure
a sustainable engagement and
contribution of the users,
managers, developers, and
facilitators in the data
ecosystem. Those key
stakeholders are incentivized to
contribute to the ecosystem in
order to foster transformation of
insights into value-adding
decisions and service provision.
The technical tools we develop are tailored to the needs,
challenges, and resources of the intended users and
beneficiaries. Additionally, our iterative cycles of co-
creation focus on data security, privacy and ethics, and
actionability.
6. 5
01 Our purpose and approach at
Dalberg Data Insights
How we use AI for good
02
8. In partnership with GiveDirectly, we built a poverty model using survey and
telecom data to identify people eligible for cash transfer programs in
Uganda
7
End Result: for each subscriber, a probability to belong to the eligible class
9. Why is this a good approach to idenify eligible people for
cash transfer programmes?
8
• Using the phone survey results, we saw a high variability in poverty
levels within districts (see figure below). Individual targeting will
improve the resource allocation when limited budget needs to be
disbursed
• Based on the model, it is possible to identify with high degree of
accuracy the poorest within a group of people
11. Mobility has decreased by ~70% under confinement
measures, but rose by 32% on Sunday due to nice weather
March 2
Phase 2
March 10
Phase 2
reinforced
March 13
Partial
confinement
March 18
Reinforced
confinement
AVERAGE NUMBER OF TRIPS OUTSIDE OF
HOME ZIP CODE, PER CAPITA
Weekend
- 69%
+ 32%
illustrative
12. Different types of mobility and different dynamics over
time
illustrative
13. While 99% of zip codes have reduced mobility, small rural
zip codes still maintain pre-confinement trips
15% OF ZIP CODES HAVE REDUCED
THEIR AVERAGE NUMBER OF TRIPS BY
<30%
-30%
-40%
-50%
Distribution of ZIP codes, ranked by evolution of average
number of trips between March 2nd and 30th, per capita
Normal travel
Half of normal
No travel
SMALL AND RURAL ZIP CODES DO
NOT SIGNIFICANTLY REDUCE
MOBILITY
Scatterplot of ZIP codes, comparing the evolution of average number of
trips between March 2nd and 30th per capita and ZIP code area size
Area size
12
illustrative
14. BRUGGE-COAST
• 152k people /
10 ZIP codes
• Risk factors:
age, inbound
mobility
KORTRIJK
• 350k people /
24 ZIP codes
• Risk factors:
work mobility
RISK SCORE
Low High
AALST-NINOVE
• 138k people /
9 ZIP codes
• Risk factors:
work mobility,
link to Brussels
HASSELT-E314
• 295k people /
19 ZIP codes
• Risk factors:
covid-19 cases
OUTER LIEGE
• 67k people /
19 ZIP codes
• Risk factors:
age, mobility
BOUILLON
• 4k people / 3 ZIP codes
• Risk factors: age, mobility mostly
unchanged, covid-19 cases
6 geographical areas are flagged as high risk of
COVID19 and should be deconfined at the later stages
illustrative