NIRIKSHAK AIOversight Inspectorate

Ministry of Statistics & Programme Implementation · MPLADS Division

Login
MPLADS • Public Expenditure Oversight

Intelligent Oversight for
Public Development Works.

Detect anomalies. Verify evidence. Strengthen accountability.

NIRIKSHAK AI assists authorised public officials and programme authorities to surface unusual patterns in MPLADS works, examine empirical supporting evidence, and prioritise cases that warrant human verification.

Learn More
Evidence-LedHuman-in-the-LoopAudit-Focused
Official Governance Portals

Dedicated Oversight Workspaces

Select your designated official authority to access dedicated surveillance, evidence logs, and review queues.

Constituency Oversight

Member of Parliament (MP)

View and track projects recommended for your parliamentary constituency.

District Implementation & Verification

District Authority

Monitor project execution, verify evidence, and log official determinations in your district.

State-Level Programme Monitoring

State Nodal Authority

Monitor MPLADS projects, cross-district progress, and compliance across your state.

Ministry of Statistics & Programme Implementation

MoSPI

Monitor MPLADS national implementation, fund flows, and cross-state analytics.

About NIRIKSHAK AI

Making public expenditure easier to examine.

NIRIKSHAK is designed as an intelligence and investigation layer for MPLADS monitoring. It does not replace official decision-making. Instead, it assists officials in surfacing risk signals, organising empirical evidence, and focusing limited inspection resources where scrutiny is most warranted.

Detection, not accusation

An anomaly score is an objective indicator for human examination—never an automatic finding of wrongdoing or fraud.

Evidence before action

Investigators review baseline peer benchmarks, duplicate clusters, and records before determining audit actions.

Core Architecture

From signal to investigation.

A structured, three-stage workflow engineered to support official audits—not substitute human judgment.

STAGE 01AI-ASSISTED DETECTION

Multi-Detector Anomaly Engine

Identifies statistical signals including cost outliers against peer groups, exact project duplicates, near-duplicate textual descriptions, and irregular allocation patterns across monitored works.

Cost BenchmarksDuplicate DetectionPattern Signals
STAGE 02EVIDENCE-LED VERIFICATION

Empirical Evidence Ledger

Brings raw project records, district & state peer medians, Integrity Passports, and baseline comparisons together so every flagged case is transparently explained with verifiable signals.

Peer MediansIntegrity PassportSignal Ledger
STAGE 03HUMAN-IN-THE-LOOP INVESTIGATION

Official Decision Support

Empowers authorised oversight authorities to conduct structured inquiries, assign physical field inspections, request contractor vouchers, and maintain immutable audit histories.

Payment GateAudit TrailOfficer Controlled
Responsible Intelligence Protocol

Every signal leads to better questions—not automated conclusions.

NIRIKSHAK AI adheres to strict public governance standards. The platform operates on four non-negotiable principles:

✓

Non-Accusatory Terminology

Cases are marked as “Flagged for review” or “Investigation priority”, never premature declarations of fraud.

✓

Disclosed Data Limitations

Unrecorded expenditure or progress fields are transparently flagged as unavailable rather than fabricated.

✓

Full Officer Audit Logging

All inquiries, document requests, and review closures are permanently logged with officer credentials.

Authorised Institutional Access

Members of Parliament, District Authorities, State Nodal Authorities, and MoSPI officials access the NIRIKSHAK platform using registered credentials and verified official roles.

Role-Based Governance & Multi-Level Audit Trail