The document discusses the Bill & Melinda Gates Foundation's collaboration with SocialCops to invest $8 million in agriculture in rural India, aiming to reduce poverty among small farmers through a data-driven approach. A data intelligence platform was created to aggregate and visualize agricultural data from 31 public sources, resulting in an interactive dashboard that informs investment strategies across various agricultural indicators. The platform aims to identify optimal areas for investment, particularly in increasing pulse production to address high infant mortality rates.

1. Data-Driven Investments
Using data intelligence to invest $8
million in agriculture in rural India
with
| case study

2. The
Problem1
Reducing extreme poverty for India’s
small and marginal farmers

3. Before we bridge the
development gap,
we must bridge the
data gap.
Melinda Gates
Bill & Melinda Gates
Foundation

6. Our
Solution2
Creating a complete data-driven
picture of agriculture in India

7. The Gates Foundation
partnered with SocialCops to
create a data-driven way for
teams at the Gates Foundation
to target their investments.
Our data intelligence platform
was deployed to aggregate
agriculture data from public
sources, clean and structure
the data, and visualize the data
in an intuitive, useful
dashboard.
Overview

8. The absence of unintended changes or errors in some data. Integrity implies
that the data is an exact copy of some original version, e.g. that it has not been
corrupted in the process of being written to, and read back from, a hard disk or
during transmission via some communications channel.
data jack (ˈdadǝ jak) n.
1. A wall-mounted or desk-mounted connector (frequently a wide telephone-style
8-pin RJ-45 ) for connecting to data cabling in a building.
Data Intelligence
data intelligence (ˈdadǝ inˈtelǝjǝns) n.
1. The process of transforming all available data — collected from the ground up,
sourced from external data sets, and extracted from elaborate internal systems —
into intelligent insights that make the best decision crystal clear.
2. The only logical way to make a decision in the twenty-first century.
data link layer (ˈdadǝ lingk ˈlāər) n.
1. Layer two, the second lowest layer in the OSI seven layer model. The data link
layer splits data into frames (see fragmentation ) for sending on the physical
layer and receives acknowledgement frames. It performs error checking and re-
transmits frames not received correctly. It provides an error-free virtual channel
to the network layer. The data link layer is split into an upper sublayer, Logical

9. Our Platform
brings the entire decision-making process to one place.
It makes even the toughest decision faster and easier.
Access
external data
Collect data
from the ground up
Connect your
internal data
Visualize data and
find insights
Transform
and clean data
• Geospatial analysis
• KPI tracking
• Geoquerying
• Strategic planning

10. Our Platform
Data from 31 different public data sources — including difficult
data on crop productivity, access to irrigation facilities, local
infrastructure, soil conditions, and more — was sourced from our
data repository.
The data was matched and aggregated in a single data set with
our entity recognition engine, which finds and corrects errors. The
data set was then transformed into district-level indices on
economy, crop productivity, female empowerment, and more.
The data and indices were visualized on an interactive
dashboard with geo-clustering, district-level comparisons,
advanced geographic queries, and detailed drill downs.
Access
Visualize
Transform

11. Our Process
1 2 3 4
Data
cleaning
Data
aggregation
Score
creation
Data
visualization

12. Funding to invest
31
External data sources
$8 million
2015-16
Years of deployment
209
Total indicators
Philanthropy
Sector involved

13. The
Story3
31 data sources, 209 indicators, and
9 indices… all in 1 dashboard

14. Data from 31 sources was
pulled from Access. The
data covered 209 indicators in 9
layers:
Economic and agricultural profile
ICT and infrastructure
Crop productivity and coverage
Horticulture productivity and coverage
Financial services
Livestock services
Nutrition
Women’s empowerment and services
Policy and advocacy
Data Aggregation
1 2 3 4
Access

15. Data Aggregation
1 2 3 4
Access
Our data goes through
extensive verification and
cleaning before being added to
Access.
Data from everywhere
Data was sourced from PDF files, web
pages, text files, images, and Excel files in
the most obscure corners of the internet.
Data triangulation
Complex algorithms were used to match
data across many disparate, inconsistent
data sets, all to zoom in on the right data
points.
Trustworthy data
Every data set was cleaned, checked for
completeness and accuracy, and
prioritized based on its relevance.

16. Data Cleaning
1 2 3 4
After all the data was
aggregated, it was cleaned
and verified on Transform.
Transform
Consistency checks
Includes intra-variable checks (checking
each variable for incorrect values) and
inter-variable checks (ensuring that data
across variables and geographies is
consistent).
Data quality assurance
All other checks needed to ensure
complete accuracy, including vertical
aggregations, missing value checks, and
external validations.
Geographic aggregation
Each data point needed to be matched with
the correct district (using a master list of
geographic standards), then all the data for
each district had to be merged into a single
data set.

17. Score Creation
1 2 3 4
For easier insights, a score
was calculated for each of
the 9 layers (crop productivity and
coverage, livestock services, etc.)
in each district.
The score provides a simple way
to…
assess the status of each
topic in a given district
quickly compare each layer
across multiple districts
Transform

18. Data Visualization
1 2 3 4
Visualize
Using Visualize, all of the
cleaned, verified data was
visualized in an interactive
dashboard with…
district comparisons
state-level view
access to raw data
intelligent query engine

19. Data Visualization
Identify clusters
for investmentVisualize
1 2 3 4

20. Data Visualization
Query and identify
focus areasVisualize
1 2 3 4
productivity of maize < 1,500 x land with assured irrigation = No x

21. Data Visualization
Zoom into any
geographyVisualize
1 2 3 4

22. Data Visualization
Multi-dimensional
crop dataVisualize
1 2 3 4
Area of the bubble shows
the relative cultivation area
for that crop, while the
color gradient shows the
crop productivity for each
district.

23. Use
Case4An example of how this dashboard
helps drive better investments

24. India has a problem with
high infant mortality rates.
Can we tackle that through agricultural investments?

25. India has a pulse problem. While the production of rice and wheat has
grown consistently since Independence, pulse production has stagnated.
This has contributed to severe undernourishment, since rice and wheat are
less nutrient-rich than pulses. By increasing pulse production, agricultural
families’ nourishment improves as they can add more pulses to their diet.
(“Pulse” refers to dried peas, beans, lentils, chickpeas, and other legumes.)
pulse
productivity
We know
that increasing
infant
mortality.
helps
decrease

26. Why are pulses less common?
Pulse cultivation requires more water than rice or wheat.
Efficient pulse production needs a
good irrigation system.

27. The takeaway: the best places to invest to
decrease infant mortality
most efficiently and quickly
are the places with…
high infant mortality
low pulse productivity
good irrigation system

28. The best place to
invest in urad (black
lentil) productivity
for improving infant
mortality

29. About
SocialCops5

30. Recognition
We’ve garnered widespread support since our start in 2013.
2015 and 2016 “40 Under 40” list
- Forbes India: 2015 “30 Under 30” list
- Forbes Asia: 2016 “30 Under 30” list
- Recognized as one of the top 10 emerging startups
by Prime Minister Modi
- Selected as one of the 35 startups to visit Silicon
Valley with Prime Minister Narendra Modi for the
India-U.S. Startup Konnect in 2015
and more…
- United Nations World Youth Summit Award
- Global Social Entrepreneurship Competition
- IBM/IEEE Smart Planet Challenge
- Singapore International Foundation
- Young Social Entrepreneurs
- Aseanpreneurs Idea Canvas

31. Press and Media
We’ve garnered widespread support since our start in 2013.
Data intelligence can be used to confront the
world’s most critical problems and make a
truly data-driven decision.
Indian Management
Tracking data that solves problems is their
mission.
Economic Times
I am thrilled with the pioneering work that
SocialCops is doing. We are limited only by
our imagination in terms of how technology
can address the challenges facing humanity.
Manoj Menon, managing director (Southeast Asia) of
Frost & Sullivan
SocialCops is taking big data in a direction
that very few companies have been able to
do: providing data and insights that can help
solve real problems for most of the planet.
Pankaj Jain, Partner at 500 Startups

32. Thank You!
For more information or to request
a demo of our platform, check out
www.socialcops.com.