top of page

Headquarter

Scienaptic AI
50 West 47th Street,

Ste. 1614, New York, NY, 10036


Ph: +1 315 609 8036

Branch offices

Florida

Texas

Maryland

  • LinkedIn
  • Youtube

contactus@scienaptic.com

© Scienaptic Systems Inc.

All rights reserved.​

Integration Guide

Typical Integration Flow

  1. Client sends historical transaction data for member accounts

  2. Cashflow Analysis processes and enriches the data

  3. Outputs are returned for consumption

  4. Client systems store and apply outputs in underwriting or monitoring workflows

Where It Fits in the Loan Lifecycle

  • Member onboarding

  • Loan application and underwriting

  • Periodic account or portfolio reviews

Cashflow Score

Description

The Cashflow Score is a predictive risk score generated using advanced machine learning models trained on cashflow attributes. It estimates the probability of default based on observed financial behavior.

Prediction Target

  • 60+ Days Past Due (DPD) in the next 12 months

Output

ATTRIBUTE
DATA TYPE
DESCRIPTION
Cashflow Score
Integer
Risk score ranging from 300–900; lower scores indicate higher risk

Interpretation

  • The score complements bureau and traditional credit signals

  • Can be used in cutoff-based, banded, or blended decision strategies

Cashflow Attributes

Description

Cashflow Attributes are aggregated metrics calculated at the account level. These attributes summarize financial behavior and liquidity trends over time and are directly consumed by underwriting rules and risk models.

Time Windows

Attributes are summarized over the following periods:

• Last 1 month
• Last 3 months
• Last 6 months
• Last 12 months

Attribute Categories

ATTRIBUTE TYPE
SUMMARIZATION
ATTRIBUTES
TIME PERIOD
SAMPLE ATTRIBUTES
Withdrawals
total, min, max, mean, variance, trend
counts and amounts of withdrawals (debits) by transaction category 
summarized over the last 1,3,6,12 months
total_debits_l3m, sd_debits_l3m, total_cash_withdrawals_l6m, recurring_debits_l3m…
Balance
total, min, max, mean, variance, trend
average daily balance 
summarized over the last 1,3,6,12 months
avg_balance_l3m, avg_balance_l6....
Cashflow
total, min, max, mean, variance, trend
monthly cashflow (credits - debits)
summarized over the last 1,3,6,12 months
cashflow_l3m, cashflow_l6m…
NSF/OD
total, min, max, mean, variance, trend
counts and amounts of NSF and OD
summarized over the last 1,3,6,12 months
nsf_count_l3m, od_fee_l3m….
Ratios
total, min, max, mean, variance, trend
balance_to_credit, debit to credit 
summarized over the last 1,3,6,12 months
balance_to_credit_l3m, debit_to_credit_l3m…
Payments
total, min, max, mean, variance, trend
counts and amounts of payments by category 
summarized over the last 1,3,6,12 months
loan_payments_count_l3m, loan_payment_amount_l1m, mobile_app_payments_l3m…
Transfers
total, min, max, mean, variance, trend
counts and amounts of transfers by category
summarized over the last 1,3,6,12 months
internal_transfers_l3m, other_transfers_l3m….
Deposits
total, min, max, mean, variance, trend
counts and amounts of deposits (credits) by transaction category
summarized over the last 1,3,6,12 months
total_credits_l3m, sd_credits_l3m, total_cash_deposits_l6m, recurring_credits_l3m…

Note: Transaction data for the last 12 months is required to generate the aggregated attributes.

Usage

​Cashflow attributes enable:

  • Fine-grained risk segmentation

  • Affordability and liquidity analysis

  • Explainable underwriting decisions

Income Estimation

Description

Income Estimation analyzes inflow transactions to infer the applicant’s earning capacity. The process distinguishes between recurring and non-recurring income sources and evaluates consistency, frequency, and historical depth.

Key Characteristics

  • Leverages categorized inflows such as payroll, government benefits, and investment income

  • Accounts for frequency and recurrence patterns

  • Assigns a confidence score based on data quality and consistency

Output

ATTRIBUTE
DATA TYPE
DESCRIPTION
Estimated Net Income
Float
Estimated yearly net (post-tax) income
Confidence Level
Integer (1–5)
Confidence in income estimate (1-5, 5 being the highest)
Estimated Gross Income
Float
Estimated yearly gross (pre-tax) income

Interpretation Guidance

  • Higher confidence levels indicate stable and recurring income patterns

  • Income estimates should be used alongside other credit and affordability signals

Transaction Categorization

Description

Transaction Categorization enriches raw transaction records with a standardized taxonomy. This enrichment enables consistent downstream analysis for income estimation, cashflow aggregation, and risk modeling.


Each transaction is classified across multiple dimensions capturing mode, recurrence, frequency, and spending intent.

Categorization Dimensions

FIELD NAME
DESCRIPTION
POSSIBLE VALUES
Inflow
An inflow is any transaction where money enters the account, increasing the account balance.

  • Payroll

  • Social Security (Pension, Disability)

  • Tax Refund

  • Interest Income

  • Card Rewards

  • Crypto

  • Bank Cash Advance

  • Loan

  • Rental Income

  • Investment Income

  • Others

Outflow
An outflow is any transaction where money leaves the account, decreasing the account balance.

  • Payments

  • Internal Transfer to Deposit Accounts

  • Withdrawal

  • Fees

  • Others

Payment Type
Describes the financial purpose or category of the transaction.

  • Savings and Investment

  • Housing Payments (Mortgage, Rent, Taxes etc)

  • Utilities (Gas, Electricity, Water, Mobile etc)

  • Transportation

  • Car Payments

  • Healthcare

  • Food and Entertainment

  • Education

  • Credit Card Payments

  • Gambling

  • Loan Payments

  • Others

Fee
Identifies transactions representing fees or penalties charged to the account.

  • NSF

  • ATM

  • Transfer Fees

  • Overdraft

  • Account Maintenance

  • Others

Inflow Source Type
Classifies the origin of incoming funds for inflow transactions.

  • Government

  • Non-Government

Mode
Describes the primary transaction mechanism used to move funds.

  • ACH

  • Wire

  • Check

  • Cash

  • Debit Card

  • Others

Sub-Mode
Provides additional channel-level detail for the transaction mode.

  • Wallet

  • Application

  • Mobile

  • Check

Recurrence
Indicates whether a transaction occurs repeatedly over time.

  • Recurring

  • Non-Recurring

Frequency
Specifies how often a recurring transaction occurs.

  • Daily

  • Weekly

  • Fortnightly

  • Monthly

  • Yearly

  • Others / NA

Transaction Type
Indicates the direction of money flow relative to the account.

  • Inflow

  • Outflow

Output Usage

Categorized transactions serve as the foundation for:


• Identifying income streams
• Detecting recurring behavior
• Aggregating category-level cashflows
• Supporting explainability in underwriting decisions

Data Model & Inputs

Cashflow Analysis requires account-level and transaction-level data. Mandatory fields must be provided for successful processing. Optional fields enhance enrichment quality and accuracy.

Account & Member Fields

FIELD NAME
DESCRIPTION
MANDATORY
Member ID
Unique identifier for the member
No
Account ID
Unique identifier for the account
Yes
Account Type
Checking, Savings, etc.
Yes
Account Ownership Type
Individual, Joint
Yes
Account Balance
Available balance on the account at time of transaction
Yes

Transaction Fields

FIELD NAME
DESCRIPTION
MANDATORY
Transaction ID
Unique identifier of the transaction
Yes
IP Address
IP address of transaction origin, if available
No
Transaction Location
Geolocation of transaction origin, if available
No
Payee Account ID
Destination account ID for transfers
No
Merchant Category Code (MCC)
Merchant category code, if available
No
Post Transaction Balance
Remaining balance on the account after the transaction
No
Transaction Currency
Currency of transaction
No
Transaction Channel
ACH, Debit Card, Branch, Online, etc.
No
Transaction Type
Deposit, Withdrawal, Payment, etc.
No
Transaction Description
Description / comments
Yes
Transaction Amount
Transaction amount
Yes
Transaction Datetime
Date and time of the transaction
Yes

Overview

Cashflow Analysis processes historical transaction data from member accounts to produce actionable insights used in underwriting, income validation, and credit risk assessment.

What Cashflow Analysis Delivers

  • Enriched and standardized transaction categorization

  • Estimated gross and net income with confidence levels

  • Aggregated cashflow attributes across multiple time windows

  • A machine-learning based cashflow score predicting default risk

High-Level Flow

  1. Transaction data is ingested at the account level

  2. Each transaction is categorized across standardized dimensions

  3. Income is inferred from inflow behavior

  4. Cashflow attributes are aggregated over time

  5. A cashflow score is generated based on derived attributes

Common Use Cases

  • Loan underwriting and credit decisioning

  • Income validation and affordability assessment

  • Risk segmentation

  • Supplementary signal alongside bureau data

Cashflow Analysis

Cashflow Analysis is a comprehensive capability that analyzes member transaction data to derive structured insights into financial behavior, income stability, liquidity, and credit risk. It transforms raw bank transactions into standardized categories, income estimates, aggregated cashflow attributes, and a predictive cashflow risk score.

For White Background.png

LendSmart AI

PreScreen AI

LimitWise AI

FraudShield AI

Early Warning Signals

Bundled X-Sell

Bundled Deposit+

bottom of page