An ISO/IEC 27001:2022 Certified Organization
Lending Software for Banks and NBFCs
The Problem it Solves
Detection has outpaced disposal. AI models find suspect accounts — then the hardest part starts, by hand, on paper, e-mail and Excel
No Consistent SOP
Every alert clears EDD through improvised branch processes — nothing repeatable, nothing observable
No Turnaround Discipline
No per-stage clocks, warnings or escalation; alerts age silently while mule funds move on
No defensible record
Review feedback is scattered in inboxes, making it impossible for inspectors to access. Regulators are now explicitly directing SOPs accounts linked to mule and cyber fraud.
Key Capabilities
01
AI Triage,structurally cited
02
Native transaction detection
Triage across 11 laundering typologies — every claim must cite a specific evidence record; the claim schema rejects uncited output. Identity resolution across accounts (CIN > PAN > mobile > e-mail)
03
STR pipeline, clock enforced
​Ingests raw EOD transaction extracts, computes windowed aggregations per customer and raises its own alerts — structuring, velocity bursts vs the customer's own baseline, dormant activation, mule fan-in. Every threshold is a policy key
04
Mule-account EDD as configured SOP
7-working-day STR clock with warning bands; sign-ready reports in FIU-IND FINnet 2.0 v1.14 formats. The platform never signs, never files — the Principal Officer does, and the loop closes on the audit trail
The bank's own SOP runs as configured stages with per-stage TAT clocks, notifications and escalation. Detection-model alerts verified independently; ring claims checked against the bank's own data
05
Dual-control enforcement
Freeze/unfreeze dispatched to the CBS under structural dual authorisation — the approve API rejects the requesting officer; idempotent dispatch can never double-freeze
06
Customer-360 at triage
Profile live from CBS/ESB; product holdings from the data warehouse with freshness disclosed; an unavailable source is recorded as a stated data gap, never a silent omission
Why SimVigil
One platform, two live configurations: the bank's STR pipeline and its mule-account EDD operation — AI proposes with structural citations, deterministic engines compute, named humans decide
01
Own findings, not just workflow
Arrives at its own detections from raw transactions — not only a case layer over other systems' alerts
02
AI never does arithmetic
Extraction proposes; versioned deterministic evaluators compute. Same inputs, same outputs, inputs on record
03
Graduated autonomy
Autonomy gates park AI-driven runs until a human releases them — the dial belongs to the bank
04
Regulatory alignment
​FIU-IND FINnet 2.0 v1.14, RBI KYC Master Direction (deemed-OVD as data), PMLA record-keeping, mule-account SOP-ready
Explore our other suite of Products
Everything you need for your lending business!