Nine Agents

Nine specialised examiners, one pass over your ledger

Each agent is scoped to one class of financial malpractice a government audit officer actually encounters, and each returns the transactions and the arithmetic behind every flag.

The Nine Agents

Nine specialised examiners, one pass over your ledger

Each agent is scoped to one class of malpractice an audit officer actually encounters. An agent only appears in the output if it found something, and every entry states the transactions and the calculation it rests on.

01Threshold evasion

BLOODHOUND

खोजी

Detects payments deliberately split to stay under sanction, tender or approval limits — repeated bills just below the delegation ceiling, round-number invoicing and serial awards to the same firm within one limit.

02Vendor collusion rings

SPIDER

जाल

Links suppliers that share a registered address, phone number, bank account, tax-ID prefix or common directors, exposing cartels bidding against themselves in the same tender.

03Ghost billing and phantom works

SKEPTIC

संशयी

Flags vague line items, works billed without measurable quantities, duplicate bills across financial years and single projects split into many small works to avoid open tendering.

04Timing anomalies

CHRONOS

कालचक्र

Surfaces March-end fund dumping to prevent lapse, sanction-to-payment completed the same day, backdated approvals and processing outside working hours.

05Geography mismatch

TRACKER

अनुचर

Compares vendor registration against the works location — suppliers hundreds of kilometres from the site, residential addresses for civil contractors, and awards outside the sanctioned jurisdiction.

06Beneficiary history

PROFILER

परिचायक

Identifies firms incorporated weeks before a large award, long-dormant entities suddenly reactivated, and beneficiaries appearing across multiple schemes under near-identical particulars.

07Head-of-account misuse

INQUISITOR

प्रश्नक

Catches expenditure booked against the wrong scheme or budget head, capital spending charged to revenue, and diversion between centrally sponsored and state-funded pools.

08Pass-through and circular routing

SPECTER

छाया

Traces value handed from a prime contractor to a single subcontractor, layered intermediaries adding no work, and funds returning to an originating entity through a chain of transfers.

09Consolidation and ranking

PROSECUTOR

अभियोजक

Reconciles the findings of the other eight into one audit brief — a single severity score per entity, the corroborating evidence, and the inspection or recovery action to consider first.

Agent reference

What each agent does, in an auditor's terms

One row per agent: the mandate it holds, the irregularity it looks for in departmental records, and the evidence it returns for the file.

  • 01Bloodhoundखोजी

    Threshold evasion

    What it flags
    Bills repeatedly cleared just under the sanction or tender ceiling, one work split into several smaller orders, and round-figure invoicing that avoids competitive bidding.
    Evidence returned
    The transaction list, each amount, the applicable ceiling, the shortfall against it and the count of repetitions in the period.
  • 02Spiderजाल

    Vendor collusion rings

    What it flags
    Suppliers that share an address, telephone number, bank account, tax-ID series or directors, so a single interest bids against itself in the same tender.
    Evidence returned
    The cluster of firms, the identifiers they share, and the tenders in which two or more of them appeared together.
  • 03Skepticसंशयी

    Ghost billing and phantom works

    What it flags
    Vague line items with no measurable quantity, the same bill raised in two financial years, and works certified without a measurement reference.
    Evidence returned
    The bill numbers, dates, amounts and the duplicate or missing particulars that led to the flag.
  • 04Chronosकालचक्र

    Timing anomalies

    What it flags
    March-end dumping to prevent lapse of grant, sanction and payment completed the same day, and approvals recorded outside working hours.
    Evidence returned
    Expenditure distribution by month and day, the sanction-to-payment interval per case, and the share of the year's spend in the final week.
  • 05Trackerअनुचर

    Jurisdiction and geography mismatch

    What it flags
    Vendors registered far from the works site, civil contractors at residential addresses, and awards made outside the sanctioned jurisdiction.
    Evidence returned
    Vendor address of record, the work location on the order, the distance between them and the sanctioning authority's jurisdiction.
  • 06Profilerपरिचायक

    Beneficiary and vendor history

    What it flags
    Firms incorporated shortly before a large award, dormant entities suddenly reactivated, and beneficiaries recurring across schemes under near-identical particulars.
    Evidence returned
    Date of registration against date of first award, the gap in transaction history, and the list of schemes the entity appears in.
  • 07Inquisitorप्रश्नक

    Head-of-account misuse

    What it flags
    Expenditure booked to the wrong scheme or budget head, capital items charged to revenue, and diversion between centrally sponsored and state pools.
    Evidence returned
    The voucher, the head it was booked under, the head the item belongs to, and the amount diverted.
  • 08Specterछाया

    Pass-through and circular routing

    What it flags
    A prime contractor handing most of the value to one subcontractor, layered intermediaries adding no work, and funds returning to an originating entity.
    Evidence returned
    The chain of payments in order, the share of value retained at each hop, and the closing loop where one exists.
  • 09Prosecutorअभियोजक

    Consolidation and ranking

    What it flags
    Nothing on its own — it reconciles the other eight so one entity flagged by several agents rises above one flagged by a single test.
    Evidence returned
    A single severity score per entity, the corroborating findings behind it, and the inspection or recovery step to consider first.

Disclosure

Operating boundary. NIRNAE holds no live access to treasury, PFMS, banking or departmental systems. The department exports the records it wants examined and submits them. The product is in pre-deployment: engagements today are supervised pilots on exported or anonymised datasets, and every finding is an indicator for verification, not a determination. Full answers in the FAQ.

Interested?

Leave a departmental email and we will arrange a briefing on a sample dataset.

Request a briefing