DPI Deep Dive — Saturday | July 11, 2026

Layer L6: Governance & Grievance — DARPG, CPGRAMS, eOffice, and the Digital Government Backbone

This week, India’s digital governance layer saw a flurry of landmark developments — from the adoption of the Jaipur Declaration at the 29th National Conference on e-Governance to the empanelment of six AI firms for government deployments, the launch of an AI-powered pension grievance portal, NIC’s first sovereign AI data centre, and sustained CPGRAMS grievance disposal volumes. Together, these signal a decisive shift from piloting AI in government to embedding it as core infrastructure for citizen services.

1. The Jaipur Declaration: India Commits to “Human-Led AI” in Digital Governance

The 29th National Conference on e-Governance (NCeG 2026), held on July 1–2 in Jaipur under the theme “Viksit Bharat 2047: AI-enabled, Data-Driven and Secure Digital Governance,” concluded with the adoption of the Jaipur Declaration — a strategic roadmap that places AI at the centre of India’s public-sector transformation, but with a critical caveat: human oversight must remain paramount. ¹

Union Minister Dr. Jitendra Singh, who presented the 17 National e-Governance Awards, framed AI as essential for governance but stressed that responsible deployment must prioritise transparency, accountability, and public service. The declaration outlines a whole-of-government framework for secure, citizen-centric, and accountable AI-enabled governance — a clear signal that India is moving beyond isolated pilots toward integrated, scaled implementation across central and state administrations. ¹²

The conference drew over 80 Union ministries and departments, 28 states, and 8 Union Territories, making it the largest NCeG gathering to date. The event also featured the launch of Rajasthan’s SMART Project (Services Management with Artificial Intelligence and Real Time System), enabling citizens to access 27 government services via WhatsApp — from utility bill payments to certificate applications — by simply sending a “Hi” to the e-Mitra WhatsApp service number. This exemplifies the kind of zero-friction, conversational-service delivery the Jaipur Declaration envisions. ³

National e-Governance Awards 2026 — Key Winners

The awards recognised 17 projects across categories, with AI-driven and digital public infrastructure initiatives dominating the Gold Awards: ²⁴

• AgriStack (Ministry of Agriculture & Farmers’ Welfare) — Gold for Government Process Re-engineering, recognising DPI-based transformation of agricultural service delivery.
• AI-Aadhaar Clinical Decision Support System (MoHFW) — Gold for Innovation through AI, assisting doctors in clinical decisions and expanding healthcare access in remote areas.
• ICMR-MINDS — Gold for integrating mental health and addiction screening with mainstream NCD care via an AI-enabled Clinical Decision Support System, deployed across seven states.
• Mahakumbh 2025 Digital Governance Initiative (Prayagraj Mela Authority) — Gold for AI-based command & control, digital mapping, and smart public services.
• Panchayat Advancement Index 2.0 (Ministry of Panchayati Raj) — Gold for data-driven assessment of 2.6 lakh gram panchayats (97.3% coverage) across 33 States/UTs.
• ISRO Space Applications Centre Cybersecurity Platform — Gold for strengthening threat detection and DNS filtering.
• CPGRAMS and SAMADHAN DIDI (BHASHINI-powered multilingual chatbot) were also highlighted as exemplars of citizen-centric digital grievance redressal.

Cross-Layer Connections

Several award-winning projects span multiple DPI layers: AgriStack (L5 Sectoral + L3 Documents), AI-Aadhaar CDSS (L1 Identity + L5 Health), and CPGRAMS (L6 Governance + L2 Payments for financial grievances). This convergence validates the stack-based approach to DPI design, where each layer amplifies the others.

2. MeitY Empanels Six AI Firms Under Government AI Mission

On July 9, the National e-Governance Division (NeGD) under MeitY empanelled six technology companies to deploy AI initiatives across central government departments, state governments, and public sector undertakings. The empanelment, selected from nearly 80 bidders in a competitive process, creates a trusted pool of AI implementation partners that ministries can engage directly — bypassing individual tenders for every AI project and significantly reducing procurement timelines. ⁵⁶

The six empanelled firms:

The empanelment covers the full AI lifecycle — strategy consulting, solution development, ML model building, intelligent document processing, citizen service automation, analytics, workflow optimization, and ongoing technical support. The framework is valid for two years, with a one-year extension option. ⁵

Why This Matters

This empanelment directly addresses one of the biggest bottlenecks in India’s digital governance: procurement delays. A typical government AI tender can take 12–18 months from concept to contract. The panel eliminates this friction for a predefined set of qualified partners. It also aligns with the IndiaAI Mission (₹10,372 crore allocation over five years) and the Cabinet Secretary’s directive for all ministries to submit AI roadmaps. NIC has already cleared 47 of 90 proposals, and NeGD has received ~700 assessments/case studies from ministries for vetting. ⁶

Execution Risk

The IndiaAI Mission has seen execution gaps — FY26 allocated ₹2,000 crore with only ~₹800 crore utilised, and the FY27 allocation was reduced to ₹1,000 crore. Whether the empanelled firms can convert framework access into actual deployments at scale remains to be seen. ⁵

3. PFRDA Launches “Pension Sahayak” — AI-Powered Grievance Portal for NPS Subscribers

The Pension Fund Regulatory and Development Authority (PFRDA) has replaced its legacy Central Grievance Management System (CGMS) with Pension Sahayak, an AI-powered grievance redressal platform for the National Pension System’s 7+ crore subscribers. The platform represents a significant upgrade in how pension-related complaints are filed, tracked, and resolved. ⁷⁸

Key features:

• 22-language support with text and voice input — critical for India’s multilingual citizen base, especially elderly and rural subscribers
• Mobile-first access via OTP login; all PRANs linked to a single mobile number appear under one dashboard
• AI triage and routing: The system automatically identifies the grievance type, categorises it, and routes it to the appropriate entity (Nodal Office, CRA, or Pension Fund)
• Real-time tracking with visibility into the handling entity and estimated resolution timeline
• Automatic escalation if actions are delayed beyond defined timelines
• End-to-end digital appeals — no more physical submissions for Ombudsman appeals
• WhatsApp integration for filing and tracking

Pension Sahayak aligns with the NextGen CPGRAMS framework, applying AI-driven workflows to a sector that handles millions of transactions monthly and where grievance backlogs have historically been a pain point. For NPS and Atal Pension Yojana (APY) subscribers, this could meaningfully reduce resolution times and improve trust in long-term pension governance.

Cross-Layer Connections

Pension Sahayak connects L6 (Governance) with L2 (Payments — NPS/APY financial flows) and L1 (Identity — Aadhaar-based authentication via PRAN). It also builds on Bhashini (L3) for multilingual voice input, demonstrating how DPI layers compose to deliver citizen services.

4. NIC Builds India’s First Government-Owned AI Data Centre in Delhi

The National Informatics Centre (NIC) is constructing India’s first government-owned AI data centre at its Shastri Park facility in New Delhi, with a capacity of 1.1 AI exaflops and 248 GPUs, expected to be operational later in 2026. ⁹¹⁰

This is a significant move toward sovereign AI compute infrastructure. Until now, government AI workloads have relied on commercial cloud providers or limited on-premise GPU clusters. The NIC data centre will provide dedicated, government-only compute for training and running AI models across central departments — keeping sensitive government data within official systems.

NIC is also developing:

• A Parliament proceedings database (for MPs, enabling AI-powered search and Q&A over legislative history)
• A legal database project that digitises and OCRs legal records for AI-driven retrieval

These databases will feed into the sovereign AI data centre, creating a closed-loop ecosystem of government data + government compute + government AI applications. ⁹

The Broader Context

India is in the midst of a massive AI data centre buildout. Google has committed ~$15 billion in Indian data centre investment, the Adani Group plans a $100 billion, 5GW AI platform by 2035, and Gujarat has launched its first Data Centre Policy targeting ₹6 lakh crore in investments and 7.5 GW capacity. The NIC centre complements these private investments by providing a secure, government-controlled anchor for sensitive workloads — a model that mirrors the “digital public infrastructure” philosophy of building government-owned backbone layers. ¹⁰¹¹

Cross-Layer Connections

A sovereign AI data centre underpins every other DPI layer — from Aadhaar authentication models to UPI fraud detection to CPGRAMS grievance classification. Without domestic compute capacity, India’s AI ambitions remain dependent on foreign cloud providers. The NIC centre is the compute analogue of UPI for payments: a government-built backbone that the private sector builds upon.

5. CPGRAMS Continues High-Volume Grievance Disposal

CPGRAMS (Centralized Public Grievance Redress and Monitoring System), the backbone of India’s digital grievance infrastructure, maintained strong disposal volumes throughout the week. DARPG’s daily reports show: ¹²¹³

These daily disposal figures — typically 15,000–20,000 total grievances resolved per day across central and state levels — underscore CPGRAMS as one of the world’s largest digital grievance platforms. DARPG noted that CPGRAMS has grown from disposing 2.85 lakh grievances in 2014 to over 22.7 million in 2025, a nearly 80x increase in a decade. ¹⁴

Meanwhile, the Department of Financial Services (DFS) secured 2nd rank in the Group A category of the Grievance Redressal Assessment & Index (GRAI) for May 2026, with its Insurance Division leading. DFS handles over 2.5 lakh grievances annually and has been in the top-10 since November 2025 — demonstrating that financial-sector governance (L2) increasingly feeds into the broader grievance redressal layer (L6). ¹⁵

PM Modi also highlighted the e-Jagriti platform this week — a unified AI-enabled system for consumer complaints launched on January 1, 2025, which consolidated four legacy systems (OCMS, e-Daakhil, NCDRC CMS, and CONFONET). e-Jagriti has processed over 2.29 lakh consumer cases with a disposal rate of 90.75%, rising to 92.3% in FY 2025-26. ¹⁶

What the Data Shows

The convergence of CPGRAMS, Pension Sahayak, e-Jagriti, and DFS’s GRAI performance represents a maturing ecosystem where grievance redressal is no longer just a complaint-box function but an AI-augmented, data-driven governance layer. Each platform feeds lessons and data back into the system, creating feedback loops that improve both service delivery and policy design.

Looking Ahead

This week’s developments in L6 Governance & Grievance reveal a clear trajectory:

1. AI governance is being institutionalised, not just piloted — from the Jaipur Declaration to the six-firm empanelment and NIC’s sovereign data centre.
2. Grievance redressal is becoming AI-native — Pension Sahayak, e-Jagriti, and CPGRAMS all demonstrate that AI classification, routing, and escalation are now table stakes, not experiments.
3. Sovereign infrastructure is catching up to ambition — the NIC AI data centre addresses the compute gap that has constrained government AI adoption.
4. Cross-layer DPI integration is deepening — every major L6 development this week connects to identity, payments, documents, or data exchange layers.

The challenge remains execution. IndiaAI Mission allocation cuts, the gap between ministry AI roadmaps and actual deployments, and the need for sustained capacity-building in civil services are risks that could slow the momentum. But the institutional scaffolding — declarations, empanelments, data centres, platforms — is now largely in place. The next 12 months will test whether this scaffolding translates into citizen-visible outcomes.

Data sources: DARPG, MeitY/NeGD, PFRDA, NIC, The Print, The Economic Times, DD India, DPI Watch archives.