NDVI based credit scoring for farmer.pdf

Team Pera Ache, Chill Nai presented this idea at Hult Prize at IUB (On Campus Round) and won 1st Runner Up position.

Core idea is to make financial inclusion more accessible for remote farmers in Bangladesh and other developing nations.

Reach out at - [email protected]

1. World’s food production need to increase by
70% by 2050

2. PROBLEMS USUALLY A
FARMER FACE

3. ACCESS TO PROPER
AND CONVENIENT
FINANCE
SOIL DEGRADATION
LAND
OWNERSHIP
WEATHER UNCERTAINTY

4. LOAN DISBURSEMENT CYCLE
Central Bank
PCBs, FCBs, SOCBs, SOSBs
NGOs
2%
9%
< 9%
16 - 25%
Farmers
Banks turn to NGOs
and microfinance
institutions for farm
loan disbursement to
lessen their supervisory
and due to low visibility
“There is no credit
product in the banking
sector for which one has
to pay an interest rate as
high as 25 percent” - SM
Moniruzzaman
Avg. credit size is
50,000 - 100,000BDT
Source: Bangladesh Bank

5. FARMERS HAVE LIMITED ACCESS TO FORMAL FINANCIAL SYSTEM
16.5 Million Number of smallholder farmers in Bangladesh
2.7 Million Number of farmers received credit from formal financial institutions
3,449 Farmers in remote areas who received credit
Why banks
are not
providing
credit to most
farmers?
1.Absence in
farmers
financial data
2.Low digital
literacy rate of
farmers
3.Documents &
other
requirements
4.Logistical and financial challenge to reach farmers in
remote areas
Source: Bangladesh Bank, The Daily Star, BRDB, BSBL

6. VISIBILITY

$
$

7. Pera ache, Chill nai
MD Rahat Al Mamun
Team Lead
Sumaiya Islam
UI/UX Designer
MD Rayhan Hossain
GIS Expert

8. Production yield estimation
using satellite imagery
ACCESS TO CREDIT

9. SOLUTION MODEL
What is NDVI?
-The Normalized Difference Vegetation Index (NDVI) is a measure of vegetation health derived from satellite or
drone imagery, calculated using the difference between near-infrared (NIR) and red light reflectance.
Identify ideal NDVI
Take historical NDVI
data
Compare both to identify the production yield
From BBS dataset
From satellite data analysis (MODIS)

10. Sown NDVI Mature Stage Harvest
Mid November March - April April - May
Rabi Season in Bangladesh
Sowing
NDVI
Mature
Harvest
NDVI
TIME

11. Calculate Coefficient of determination using ground truth data
Yield estimation is set to become a game-changing tool for financial institutions and agri-fintech
companies. By leveraging data-driven insights, it enables smarter decision-making—assessing a
farmer's track record, evaluating land performance, and predicting the likelihood of success.

12. TARGET GROUP
Formal Financial Institutions Agricultural Fintech

14. FINANCIAL BREAKDOWN
Cost
Business Model
Year Hectares Monitored Revenue COGS Gross Profit
1 1000 320000 145000 -500000
2 2000 640000 29000 -200000
3 3000 960000 435000 525000
4 4000 1280000580000 700000
Unit cost per hectare: EOS Crop Monitoring Dataset (20BDT) + Processing cost (70BDT) + Storage cost (5BDT) +
Labour cost (30BDT) + Operational cost (20BDT)
Tiered Service Subscription Based
Data Analytics
Service
Training and SupportAPI Access

15. CAPTURING MARKET
Entry Positioning Scale
●Collaboration with
agri fintech startups
such as - iFarmer
●Conducting small
pilot projects
●Gathering dataset
●Launch business models
●Capture contract farmers
●Tie with banks optimizing
there supervisory cost for
remote credit
disbursement
●Add more features to
the product
●Create E-KYC Platform
for more robust
profiling
●Building proprietary in
house solution

16. TARGET SDGs

17. THANK
YOU

18. Appendix
EOS Crop Monitoring cost

19. MongoDB Atlas Costing