DPI Deep Dive — Friday | August 07, 2026

L5: Sectoral Infrastructure — ABHA, AgriStack, eCourts

This week’s Sectoral Infrastructure layer saw decisive action on three fronts: AgriStack’s Digital Crop Survey preparations intensified across multiple states, eCourts Phase III pushed deeper into paperless court operations with new digitisation mandates, and ABHA crossed the 94-crore mark even as the utilisation gap in digital health records widened.

1. Odisha AgriStack SLSC: 20.75 Lakh Farmers Registered, Kharif Digital Crop Survey from August 16

The 8th State Level Steering Committee (SLSC) of AgriStack-Odisha, chaired by Chief Secretary Anu Garg on August 4, reviewed accelerating farmer registration and finalised preparations for the Kharif 2026 Digital Crop Survey (DCS).¹²

Key takeaways:

• Farmer registration hits 20.75 lakh — a meaningful number but still a fraction of Odisha’s estimated 50+ lakh farmer households. The state is targeting faster onboarding through block-level facilitation camps.
• Rabi 2025 DCS post-mortem: 53.78 lakh hectares were surveyed, but crops were identified on only 19.91 lakh hectares — a 37% identification rate that raises questions about ground-truth accuracy, training adequacy, and data quality workflows.
• Kharif 2026 DCS commences August 16, running 45 days to two months. Field staff training at block and district levels is scheduled August 5–17. Survey protocols mandate data collection approximately one month after sowing and at least 15 days before harvest — a tighter window designed to improve crop identification accuracy.
• The SLSC also reviewed the AgriStack farmer registry’s linkage to land records and demographic data, a prerequisite for the Farmer ID (Kisan Ki Pehchaan) that will serve as the farmer’s digital identity across credit, insurance, and procurement services.

Cross-layer connection: AgriStack’s Farmer ID architecture mirrors Aadhaar (L1) in design philosophy — a federated, consent-driven identity layer. However, unlike Aadhaar’s statutory backing under the Aadhaar Act, AgriStack lacks dedicated legislation, leaving data governance to MoUs between the Centre and states.

Why it matters: Odisha is among the more active AgriStack states. If its DCS accuracy improves this Kharif season, it could serve as a replicable template for other states still struggling with the basics of plot-level crop enumeration. The 37% Rabi identification rate, however, is a caution signal — scaling fast without improving quality risks creating a Crop Sown Registry that insurers and policymakers cannot trust.

2. AgroSpectrum Technovate 2026: Industry Lines Up Behind AgriStack as India’s “UPI Moment” for Agriculture

At AgroSpectrum Technovate 2026 (August 6–8, Greater Noida), the dominant theme was AgriStack’s role as the digital backbone for Indian agriculture. Policymakers, technology providers, and agri-market innovators discussed how digital public infrastructure could reshape agricultural decision-making and value chains.³

Notable developments from the event:

• “UPI moment” framing: Multiple speakers described AgriStack as agriculture’s equivalent of UPI — an open, interoperable platform that lowers transaction costs and enables innovation at the edges. The analogy is aspirational but telling: UPI succeeded because NPCI built the rails and let the private sector build the apps. AgriStack’s architecture is similar (central registries + open APIs), but adoption remains supply-driven rather than market-pulled.
• Digital Crop Survey scale-up: The government is scaling up DCS across the current Kharif season, with multiple states including Gujarat (Dahod district), Odisha, and others commencing surveys between August 10–16. Sikkim’s District Collector (Namchi) issued notices on August 1 for AgriStack Farmer Registry camps under the Digital Agriculture Mission.
• AI-enabled advisory layer: Discussions highlighted the emerging stack above AgriStack — AI advisory tools like Digital Green’s FarmerChat (which crossed 10 lakh users this week, covered separately below) that consume DCS data to deliver hyperlocal crop recommendations.
• Data governance concerns: Private sector participants flagged limited API access, opaque approval processes, and non-standardised data-sharing agreements as barriers to building production-grade applications on AgriStack. One speaker described the current state as “policy-ready but production-immature.”

Cross-layer connection: AgriStack’s UPI analogy extends to its relationship with ONDC (L4) — just as ONDC aims to commoditise commerce discovery, AgriStack aims to commoditise farmer identity and crop data, enabling any service provider to plug in. The difference is that ONDC has relatively standardised product catalogs, while crop data is inherently messy and seasonal.

Why it matters: Technovate 2026 signals that the industry is bought into the AgriStack vision. But the gap between conference enthusiasm and field-level reality remains wide. Until the Farmer ID and Crop Sown Registry reach production-grade quality at scale, AgriStack will remain an impressive architecture in search of reliable data.

3. FarmerChat 2.0: AI Advisory Crosses 10 Lakh Users, Targets 45 Lakh by 2028

Digital Green India announced that FarmerChat, its multilingual AI-powered farming advisory assistant, crossed 10 lakh users since its October 2024 launch, while unveiling FarmerChat 2.0 — an upgraded version designed for deeper personalisation and offline-first access.⁴⁵

The numbers:

• 10+ lakh users across five languages, with over 30 lakh queries answered.
• 45% women users — a significant gender balance achievement in a domain where digital tools typically skew male.
• Available in low-connectivity environments, answering questions across crop planning, pest management, livestock care, and weather forecasts.
• Digital Green targets 35 lakh additional users by 2028, focusing on women farmers, climate-smart practices, and income improvement.

FarmerChat 2.0 improvements:

• Hyperlocal advisory based on the farmer’s location, crop calendar, and historical data.
• Better integration with government databases (potentially including AgriStack’s Crop Sown Registry once APIs are open).
• Enhanced voice and vernacular capabilities for farmers with limited literacy.

Cross-layer connection: FarmerChat sits atop the AgriStack stack — it is the “last mile” application that converts infrastructure data into actionable advice. Its success depends on AgriStack delivering reliable crop and soil data. Without that, FarmerChat risks becoming a sophisticated chatbot disconnected from ground reality.

Why it matters: 10 lakh users in under two years is a genuine milestone for agricultural AI in India. The 45% women participation rate is particularly notable — it suggests that voice-first, vernacular AI can bridge the gender digital divide in agriculture. But the real test is whether FarmerChat’s advice actually improves yields and incomes at scale. That requires rigorous impact evaluation, not just user counts.

4. eCourts Phase III: Paperless Mandate Expands — Andhra Pradesh, Gauhati High Court, and 660 Crore Pages Digitised

Multiple High Courts advanced their paperless court operations this week under eCourts Phase III (₹7,210 crore, 2023–2027), with several issuing concrete operational directives.⁶⁷

Key developments:

• Andhra Pradesh High Court issued a notification fixing August 1, 2026 as the effective date for scanning and digitisation of all newly instituted cases across its jurisdiction. This is a significant operational mandate — it means every new filing enters the system digitally from day one, rather than being digitised retroactively. The notification covers all subordinate courts under the APHC’s administrative control.
• Gauhati High Court issued Notification No. 49 dated August 5, 2026, detailing “arrangements to streamline, supervise and expedite the digitization process for implementation of the Paperless Court” system. The notification establishes institutional mechanisms for monitoring digitisation progress across the Northeast’s courts.
• National scale: Digital India’s social media channels highlighted that eCourts Phase III now covers data from 4.9 crore High Court cases with over 660 crore pages digitised and 1.07 crore cases filed online. The National Judicial Data Grid (NJDG) provides real-time case statistics from 18,735 district courts and all High Courts, with over 32 crore orders/judgments accessible.
• Ministry of Law and Justice ran an active social media campaign (August 3–6) promoting eFiling, Virtual Courts, and the eCourts Services App, encouraging lawyers and litigants to adopt digital filing and online case tracking.

AI in courts — SUPACE and Adalat AI:

Digital India and IndiaAI also highlighted AI tools being piloted in the judicial system:

• SUPACE (Supreme Court Portal for Court Efficiency) assists judges with data collection and analysis for information access (not judicial decision-making).
• Adalat AI provides AI-powered real-time transcription of court proceedings, reducing the need for manual documentation.
• The Supreme Court has constituted an Artificial Intelligence Committee to explore AI’s role in the judicial domain, with an emphasis on maintaining judicial independence and cognitive autonomy.

Cross-layer connection: eCourts Phase III intersects with Aadhaar (L1) for lawyer and litigant authentication, DigiLocker (L3) for document submission, and the broader Digital India infrastructure for connectivity. The paperless court mandate also strengthens the case for DPDP Act (L7) compliance, as courts handle sensitive personal data that must be protected under the new data protection framework.

Why it matters: The APHC’s August 1 mandate is the kind of specific, enforceable directive that Phase III needs more of. Too often, eCourts announcements remain aspirational. Fixing a commencement date for mandatory scanning of new cases is operational governance in action. The 660-crore-page digitisation figure is staggering — but the real metric to watch is what percentage of those pages are machine-readable (OCR’d) and searchable, versus mere image scans.

5. ABHA Crosses 94.87 Crore IDs — But Utilisation Gap Persists

The Ayushman Bharat Digital Mission (ABDM) continues to add ABHA IDs at pace, with the latest figures showing 94.87 crore ABHA IDs created as of July 20, 2026, along with 100+ crore health records linked and 5.36 lakh health facilities registered.⁸⁹

The scale is undeniable, but a peer-reviewed study published this week reveals the utilisation gap:

• Approximately 26,000 facilities actively transact on ABDM, mostly government facilities (~21,000). Private-sector participation remains limited (~4,000).
• About 1.1 lakh facilities actively link health records to ABHA, out of ~2.5 lakh with ABDM-enabled software — meaning over half of ABDM-enabled facilities are not actively using it.
• A cross-sectional study of OPD attendees found that while awareness of ABHA is substantial (~60%), active engagement remains around 10%. Patients often create ABHA accounts at the hospital counter but rarely revisit the app, link records, or use consent flows meaningfully.
• The study describes a risk of ABHA becoming a “non-dynamic identity layer” — a credential that exists in databases but doesn’t facilitate actual health data portability in clinical workflows.

Equity progress:

A longitudinal analysis (January 2024 – July 2025) found that ABHA saturation rose from 32.7% to 53.9% nationally, with a catch-up effect in poorer states (145.2% expansion in the poorest quartile vs. 58.4% in the richest). The Concentration Index shifted from pro-rich (+0.062) to near-equitable (-0.011), driven largely by supply-side, frontline-assisted onboarding by health workers in rural areas.

Industry response:

• Hospital management software providers are increasingly building ABDM compliance into their products, with the unified health interface (UHI) enabling appointment booking, teleconsultation, and diagnostic test booking across registered providers.
• The National Health Claims Exchange (NHCX) is working to standardise insurance pre-authorisation and claim filing using ABHA-linked records, though adoption remains early-stage.

Cross-layer connection: ABHA’s authentication is built on Aadhaar (L1), and its document exchange protocols intersect with DigiLocker (L3). The private-sector utilisation gap mirrors a broader DPI challenge: building infrastructure is necessary but not sufficient. Without genuine workflow integration and demand-side incentives, digital identity layers risk becoming passive databases.

Why it matters: India now has the world’s largest digital health ID system. The equity gains are real and commendable. But the utilisation data tells a more complex story — 94 crore IDs do not equal 94 crore active users. The next phase of ABDM must shift from “create IDs” to “create value” for patients and providers. Otherwise, the 100-crore health records milestone risks becoming a statistical achievement rather than a clinical one.

Cross-Layer Snapshot

Quality Checklist:

• Cover full 7-day window (July 31 – August 7, 2026)
• 5 substantive stories with analysis
• Cross-layer connections included
• Official and primary sources prioritized
• GitHub push pending