Each topic below explains the evidence, the business problem it creates, how GLOBEIR solves it, the benefits to your organisation, and how your data stays private and secure.
1. Branch, ATM and Correspondent Network Planning
How it works. Drive-time or distance catchments are drawn around existing outlets and overlaid with village and census population to reveal white space: unserved or under-served areas. Candidate sites are scored on population, existing deposits and credit, competitor density and road access, producing a ranked shortlist for branches, ATMs or banking-correspondent points.
Proven examples.
- Jan Dhan Darshak, India. A GIS-based app developed by the Department of Financial Services and NIC that locates banking touch points across all providers. Its policy aim is a banking outlet within 5 km of every inhabited village. Banks upload the GIS locations of their branches, correspondents and ATMs. As of September 2020 it held 1.66 lakh branches, 4.35 lakh banking correspondents and 2.07 lakh ATMs, and 5.52 lakh of 5.53 lakh mapped villages (99.8%) had a branch or correspondent within 5 km [11][12].
- PMJDY Sub-Service Areas. Banks mapped all villages into 1.59 lakh Sub-Service Areas, each serving 1,000–1,500 households, so that every area has at least one fixed-point banking outlet [12][13].
- Regulatory driver. Under RBI's 2017 branch authorisation policy, at least 25% of banking outlets opened in a year must be in unbanked rural centres [12].
- FSPmaps (Gates Foundation / CGAP). Combining geo-referenced access points with population data showed that in Nigeria nearly 44 million people lived in areas with mobile coverage but more than 5 km from a financial service point. The central bank used the insight to track its inclusion strategy [14].
Business problem
Outlet coverage is now near-universal (99.8% of mapped villages have a branch or correspondent within 5 km) [12], so the question has shifted from "where are we missing?" to "where are we over-built, under-used or losing business?". Each new branch is a long-term fixed cost, overlapping catchments cannibalise one another, and the bank must still place at least 25% of new outlets in unbanked rural centres [12]. Decisions taken on spreadsheets and local opinion make it hard to defend a site, a merger or a closure to the board.
Our solution
- Geocoding of every branch, ATM and banking-correspondent point, plus competitor and Jan Dhan Darshak-style outlet layers.
- Drive-time catchments overlaid with village and census population, deposits and credit per catchment, and road access.
- White-space and site-suitability scoring that produces a ranked shortlist for new outlets, and overlap analysis that flags candidates for merger or relocation.
- Branch-performance heatmaps and a WebGIS portal for network and regional teams; rural-centre tagging to track the 25% requirement.
Benefits
- Capital spent on outlets with a measurable catchment, not on guesswork.
- Clear evidence for merging or relocating overlapping or under-used outlets.
- Easier tracking of the 25% unbanked rural-centre requirement [12].
- A shared map for network, business and compliance teams instead of separate spreadsheets.
- Industry example: geo-referenced access-point mapping helped a central bank track its financial-inclusion strategy [14].
Privacy & data security
- Network planning needs outlet locations, aggregated deposits and credit, and population layers. Customer names and account numbers are not required: analysis runs on aggregated or pseudonymised data at catchment or village level.
- Branch performance data is commercially sensitive. It is protected by role-based access (branch, region and head-office roles) and encrypted in transit (TLS 1.2 or higher) and at rest (AES-256).
- Data is hosted in India on ISO 27001 / SOC 2-certified cloud infrastructure, or in the bank's own data centre, in line with RBI outsourcing requirements that data be stored only in India [15][16].
2. Agri and Rural Credit with Satellite Data
In August 2020 ICICI Bank began using satellite imagery to assess the creditworthiness of farm borrowers. Reports covering 40+ parameters on land, irrigation and crop patterns are combined with demographic and financial data. Because land verification is contactless, credit assessment takes a few days, compared with up to 15 days under usual practice. The service started in 500+ villages, with plans to scale to over 63,000 [17]. Research with the University of Liverpool explains why this matters: 87% of Indian agricultural households are small and marginal farmers, and a large share of agricultural borrowing still comes from informal sources [18].
Rural property is also becoming bankable. SVAMITVA has prepared more than 2.42 crore property cards from drone surveys, and the cards help owners use their property for bank loans [19].
Business problem
Agri and rural loans need a physical check of the plot, the crop and the irrigation source. Manual field verification takes up to 15 days [17], adds travel cost to small-ticket loans and is hard to audit later. Without independent land and crop evidence, banks either lend slowly or lend with less certainty, and lose ground to lenders who already use satellite data [17]. Much of the target market, small and marginal farmers, still borrows informally [18].
Our solution
- Plot-level satellite indicators for the applicant's land: cropped area, crop condition over the season, irrigation signals and land-use history.
- Overlay with village boundaries, SVAMITVA-type property data where available [19], and flood and drought layers for the plot's location.
- A short, standard plot report attached to the loan file, and seasonal re-checks for monitoring and renewals.
- Field confirmation, where needed, through geotagged visits in the My GLOBEIR app.
Benefits
- Faster, contactless land and crop verification. Industry example: ICICI Bank cut agri-credit assessment from up to 15 days to a few days [17].
- Lower verification cost per small-ticket loan.
- An objective, repeatable record of the plot that auditors can review.
- Better monitoring through the season and at renewal, not only at sanction.
- A route to serve more small and marginal farmers through formal credit [18].
Privacy & data security
- The assessment is location- and land-based: it describes the plot, its crop and its hazards, not the personal characteristics of the borrower.
- Plot boundaries linked to a loan are personal data under the DPDP Act [20]. GLOBEIR processes them only for the bank's agreed purpose, under a contract, and does not sell, share or reuse them to train models for other clients.
- Where the bank lends through digital channels, the Digital Lending Directions require need-based data collection with prior explicit consent and an audit trail, and storage on servers in India [21]; the workflow collects only the plot location needed for the assessment.
3. Location-Based KYC and Fraud Prevention
Location is now built into India's KYC rules. Under RBI's KYC Master Direction, video-KYC recordings must carry the customer's live GPS coordinates and a date-time stamp, and connections from IP addresses outside India or spoofed IPs must be blocked. For digital KYC, the customer photo must be watermarked with the GPS coordinates, form number and the official's details [22]. India Post's DIGIPIN adds a precise grid-based address (about 4 m × 4 m) that banks can use as an additional KYC attribute [23].
Business problem
GPS coordinates are mandatory in video-KYC and digital KYC [22], yet many banks store them only as text inside a recording or a photo. They are not checked against the declared address, so a mismatch or a cluster of applications from one point goes unnoticed until an audit or a fraud loss. Correspondent and field staff capture KYC on many devices, which makes it hard to prove when and where each record was created.
Our solution
- Extraction of the GPS coordinates and timestamps already captured in V-CIP and digital KYC, and comparison with the geocoded declared address (and DIGIPIN where provided [23]).
- Exception rules for distance mismatches, many applications from the same point, or capture outside the expected service area.
- Geotagged, time-stamped capture for assisted KYC through the My GLOBEIR app, with one-time location access at onboarding.
- Exception maps and reports for fraud-control and audit teams on the WebGIS portal.
Benefits
- Uses location data the bank already has to meet the KYC Master Direction [22].
- Earlier detection of suspicious patterns before accounts are opened or loans disbursed.
- A clear, auditable trail of where and when each KYC was captured.
- Fewer manual re-verifications, because checks are targeted at exceptions.
Privacy & data security
- KYC data (identity, photo, GPS coordinates) is among the most sensitive data a bank holds. Analysis can run on pseudonymised records, with identities resolved only inside the bank's own systems.
- In line with the Digital Lending Directions, app access to location and camera is one-time, for onboarding and KYC, with explicit consent, and biometric data is not stored [21].
- Access is limited to named roles on a need-to-know basis, as RBI outsourcing rules require [15][16], and every view and change is recorded in audit logs.
4. Collateral and Property Verification
Lenders geocode the collateral address, check it against land-record and satellite layers (built-up area, flood zone, encroachment), and capture geotagged, time-stamped photos at sanction and at end-use verification. This creates a visual, auditable record of what was financed and where.
Business problem
Collateral and end-use checks are often a site visit, a few photos and a written note. Photos can be reused, addresses can be vague, and nothing proves that the valuer or officer actually stood at the property. Weak evidence leads to audit observations, disputes at recovery and exposure to properties in flood zones or with encroachment issues that were never mapped.
Our solution
- Geocoding of every collateral address and overlay with satellite imagery, land-record layers where available, and flood and other hazard layers.
- Geotagged, time-stamped photo capture at sanction and end-use verification through the My GLOBEIR app, with location-accuracy and duplicate-photo checks.
- A collateral map for credit, legal and audit teams, with drill-down to the visit history of each property.
Benefits
- Auditable proof of each visit: where, when and by whom.
- Detection of reused photos and visits recorded away from the property.
- Hazard flags (for example, flood zone) visible before sanction.
- Fewer audit observations related to field verification.
- A ready, location-accurate collateral base for climate-risk work (topic 5).
Privacy & data security
- Collateral addresses, property photos and visit trails are personal and confidential data. They are encrypted in transit (TLS 1.2 or higher) and at rest (AES-256).
- Field officers' location is captured only for the visit as a work record; background tracking is optional, opt-in and shown with a persistent notification.
- Offline captures are stored on the device and synced when connectivity returns, then retained or deleted according to the bank's policy.
5. Climate Risk and Disclosure
- RBI draft Disclosure Framework on Climate-related Financial Risks (Feb 2024). It covers scheduled commercial banks, large urban co-operative banks, all-India financial institutions and upper-layer NBFCs. It is built on four pillars (Governance, Strategy, Risk Management, and Metrics & Targets), with the first disclosures proposed from FY2025-26, and covers physical and transition risk, scenario analysis and geographic exposure [5].
- NGFS (the central banks' Network for Greening the Financial System) highlights gaps in physical-risk data and the options for exploiting geospatial tools [8].
- In practice: every loan and collateral asset must be accurately geocoded before a bank can say what share of its book lies in flood, cyclone or drought zones.
Business problem
The RBI draft framework expects banks to report geographic exposure to physical climate risk [5], but most loan books hold addresses, not coordinates. Without accurate geocoding, a bank cannot say how much of its portfolio sits in a flood plain or a drought-prone district, and physical-risk data itself has known gaps [8]. Building this capability late means rushed, low-quality disclosures.
Our solution
- Geocoding of loans and collateral to the best available accuracy, with a quality flag for each record.
- Overlay with flood, cyclone, drought and heat layers to assign a hazard exposure to each location.
- Exposure tables and maps by district, sector and product, structured for disclosure work under the four pillars [5].
- Scheduled refresh as the portfolio and hazard data change.
Benefits
- Readiness for climate-related disclosure as the framework is finalised [5].
- A clear view of concentration in hazard zones for risk appetite and pricing decisions.
- Better preparedness for disaster events, including where to focus relief or restructuring.
- One geocoded loan book that serves network, credit, audit and climate teams.
Privacy & data security
- Exposure is measured by location and hazard, not by borrower characteristics, and reported in aggregate by district, sector and product.
- Account-level identifiers are not needed for hazard overlay; GLOBEIR works on masked or pseudonymised loan IDs that only the bank can link back.
- Results and source data stay in India and are returned to the bank, then deleted at the end of the engagement or on request.