For Inward Cheque Clearing

Automates Inward Clearing Cheque processing using Cheque Truncation Images received from Automated Clearing House with Best in Class Automation at Various Stages

VALIDATION BY AUTOMATED
VISUAL CHEQUE SCRUTINY

ROBOTIC SIGNATURE
VERIFICATION

SIGNATURE MANDATE RULE
VALIDATION

KEY INFORMATION
EXTRACTION

EXTRACTS HANDWRITTEN &
PRINTED ENGLISH TEXTS

ROBOTIC VERIFICATION & VALIDATIONS

  • Powered by Patterns 3rd generation Digital Bot Workers[DBW]
  • Recognizes & Verifies signatures including validations against signing power mandates
  • Extracts, Transforms & Interprets printed and handwritten text information using Deep Learning Robotic elements/Patterns Defintion Language PDL’s equipped with Teacher Bots specifically trained for Cheques
  • Performs authenticity checks based on – Logo, MICR Band, Endorsement Stamp, Account Payee Stamp, Crossed Cheques
  • Identifies strikeouts, overwriting, damage, mutilation, white ink/oil/water marks on Cheques
  • Detects absence of mandatory information like date, signature etc.
  • Equipped with configurable validation engine capable of performing technical validations required for an Inward Clearing Cheque Scrutiny
  • Every Robotic Extraction, Verification and Validation is supported by justification and audit trail logging

 

Dexter Checker Queue (DCQ)

Supports incremental manual processing of failed Robotic validations and also provides conservative manual checker scrutiny for High Value / High Risk Cheques

Unknown Language Queue (ULQ)

Supports manual processing of non-English language cheques

Dexter Guided Queue (DGQ)

Provides for continuous On-The-Job Feedback training to robotic elements within Patterns Dexter

Quality Control Queue (QCQ)

Provides random selection and human quality control check of cheques automatically cleared by Dexter

INNOVATIONS

  • Symbol Identification Process – To recognize symbols specifically used by customers while writing their cheques
  • SEAM – Signature Extraction and Matching Algorithm
  • Traversal Algorithm – To separate the overlying objects in cheques
  • Cheque Approved Signatures (CAS) enables a mechanism of continuously replenishing customer signatures to record variations over time, facilitating tracking and automated addition of new specimen signatures
  • Audit Trail & Justification Log – Maintains log for all robotic decisions including the verification of signatures (Signature Justification Log Images)
  • Dexter Robotic Call Back – To secure customer confirmation before execution of a payment transaction
  • Human Robot Collaboration –  seamless means for human-robot collaboration that enables automatic feedback and bot-learning
  • Bot Worker Confidence – complete control to operations users where they could define the % confidence for each of the 30+ validations performed by the bot.

VALUE DELIVERED

  • 15 million cheques processed. All cheques are scrutinized, reduced customer complaints by 95%.
  • Seasonal High volume period volumes handled without the addition of human power.
  • Reduced human work share from 100% to 30%. 70% of the work is done by software bots.
  • The audit department is presented with a complete audit trail. Reduces audit time by over 85%.
  • 100% adherence to Turn around time.

Icons used from Collection available on www.icons8.com