AU2021104138A4 - Automated real-time digital mortgage application and decisioning engine - Google Patents

Abstract

A platform server (100) serves a web app (20) via a network (200) to a user (10) receives property loan application data at the platform server (100) from the web app (20) concerning a property loan application made by the user (10), the property loan application data comprising at least a monetary loan amount, and a property location. The automated decision engine calculates whether or not the property loan application meets one of multiple threshold requirements for automated approval of the property loan application by applying predetermined logic rules recorded at the platform server (100) using the property loan application data and data requested from a property valuation server (210), credit report server (230), financial identification server (250), and banking data server (250). FIG 1 17 1/8 z 0 U0 z 4 0 or UJ oi i z 0 J. z- 0< ) < 0 m CD tz > 0 or 01

Description

1/8

z 0 U0 z 4

0

or UJ

oi i z

J. z- 0< 0 )

<

m CD

tz > 0 or data from external data services and automates a decision-making process. This process in relation to a property loan application represents an inventive contribution in the art having various advantages, as are manifest.

[0010] Particular specific advantages comprise speed, convenience and transparency of service to loan applicants. An operator of the property loan application service is able to achieve efficiencies not otherwise possible, such as a faster, more efficient, transparent, reliable, and consistent service. Such a process is also better able to comply with credit policies, governance obligations, and regulatory requirements through systemic auditability and monitoring.

[0011] Financial services such as property loan application services have to date defied automation owing to the complexity of the process as well as the received view that an element of human judgement and management is considered essential to implement sound lending practices, as well as meet regulatory obligations and provide good communication with a loan applicant.

[0012] A preferred embodiment of the invention comprises interaction with multiple data servers, namely a property valuation server, credit report server, decision engine server, financial identification server and banking data server. Each of these servers provide automated data in response to calls from a platform server that reference selected data items from a property loan application data record.

[0013] The platform server creates and progressively builds a property loan application record that is processed according to a series of predetermined logic rules or calculations that check whether or not the property loan application meets various threshold requirements that are deemed necessary to result in automated approval of a property loan application.

[0014] Threshold requirements are checked by applying predetermined logic rules to automated data responses and property loan application data and thus calculating at the platform server whether or not the property loan application meets one of multiple threshold requirements for automated approval of the property loan application.

[0015] As will be appreciated, the predetermined logic rules implemented in a preferred embodiment are a matter of implementational detail selected at the discretion of a person skilled in the art. The predetermined logic rules are inherently arbitrary in nature and continually adapted as necessary to respond to prevailing business conditions, regulatory requirements and sound lending practices.

[0016] Additional threshold requirements not described in connection with the preferred embodiment may be implemented according to variations in credit policies, such as for example, restrictions upon loan amount, or minimum or maximum land area or floor space of a property.

[0017] The automated nature of the property loan application service described herein does not preclude supplemental use of manual intervention as circumstances dictate. This may be in the event of property loan applications that fail to meet threshold requirements, or where the predetermined logic rules fail to contemplate atypical circumstances. Manual intervention so far as it may be implemented is not specifically described as it is not of relevance to describing the automated structure and operation of the automated decision engine of the present application.

[0018] According to one aspect of the present invention there is provided a computer-implemented method for providing financial services comprising:

[0019] receiving user-reported property loan application data at a platform server, via a network from a web app accessed by a user using a personal computing device, and storing at the platform server the user-reported property loan application data as part of a property loan application record;

[0020] transmitting from the platform server to one or more data servers external to the platform server at least one data request based upon one or more data items selected from the property loan application record, and receiving in response from the data server an automated data response at the platform server;

[0021] storing the automated response data at the platform server as part of the property loan application record; and

[0022] calculating at the platform server whether or not the property loan application represented in the property loan application record meets one of multiple threshold requirements for automated approval of the property loan application by applying predetermined logic rules recorded at the platform server to one or more data items selected from the property loan application data record.

[0023] As will be evident to those skilled in the art, the preferred embodiment may be varied in terms of the data services consulted, the sequence of processing steps required, and well as the particular details of predetermined logic rules and associated parameters that are collectively used to implement the automated decision engine as described and depicted herein.

BRIEF DESCRIPTION OF THE DRAWINGS

[0024] FIG 1 is a diagram depicting in schematic form elements of the physical infrastructure of an automated decision engine according to an embodiment of the present invention.

[0025] FIG 2 is a diagram depicting in schematic form a computer system as depicted in connection with FIG 1 when implemented as a desktop computer.

[0026] FIG 3 is a diagram depicting in schematic form a smartphone as depicted in connection with the automated decision engine of FIG 1.

[0027] FIG 4 is a diagram depicting in schematic form a logical architecture of an automated decision engine implemented using the physical infrastructure of FIG 1.

[0028] FIG 5 is a flowchart of steps involved in operation of the automated decision engine depicted in FIG 4.

[0029] FIGS 6A-6C are collectively a sequence diagram representing the steps of FIG 5, as implemented in the architecture of FIG 4.

DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS

[0030] FIG 1 depicts physical infrastructure associated with an automated decision engine of a preferred embodiment of the present invention.

[0031] A platform server 100, also for convenience referred to as simply server 100, communicates via network 200 to a web app 20 through a suitable personal computing device (such as computer 70, tablet 80 or smartphone 90). Network 200 comprises the public Internet (for convenience these terms are used interchangeably), access to which is provided by various internet services providers.

[0032] An app 20 is preferably delivered through a browser (not shown) installed on the personal computing device 70-90 by way of non-limiting example. As an alternative, the app 20 may appear on the device as an app native to an operating system of the personal computing device 70-90. The device 70-90 is not limited to the examples indicated but may conceivably embrace other hardware or interfaces.

[0033] Platform server 100 may be variously implemented as cloud-native, cloud hosted or on-premise as a matter of preference.

[0034] A loan applicant 10 (otherwise termed user 10) uses the device 70-90 to engage with the app 20 as a means of using the service provided by the automated decision engine, as described below in further detail.

[0035] Server 100 is supported by various ancillary servers as indicated, with which server 100 communicates via network 200. These are as depicted property valuation server 210, credit report server 230, decision engine server 240, financial identification server 250, banking data server 260. Banking data server 260 draws on banking data sourced from various bank servers 270.

[0036] As depicted, each of these servers 210-270 is in communication with server 100 and is representative of third-party services offering data on a commercial or non-commercial basis. One or more of these servers 210-270 alternatively may be selectively incorporated into server 100 as required.

[0037] Platform server 100 comprises backend server 110, which embraces backend logic module 120, and database 130. The backend server 110 is in communication with database 130.

[0038] Each of the servers 210-270 communicates with platform server 100. Notably, each of the servers 210-270 may be configured to communicate with each other in some implementations. According to a preferred embodiment described herein, banking data server 260 communicates with bank server 270, and is used by platform server 100 to obtain data from bank server 270 without direct call to bank server 270.

[0039] FIG 2 is a diagram depicting a high-level schematic of a computer system of the type that may be used in connection with the web app 20 described above in connection with FIG 1 and referenced in the description that follows.

[0040] The representation presented abstracts various physical-layer hardware details, which may be variously implemented to provide required functionality.

[0041] Computer system 70 comprises a central processing unit 720 (CPU) that connects via a system bus (not shown) to a storage 730. Storage 730 comprises volatile memory 732, and non-volatile memory 734, each of which interoperates under direction of the CPU 720. The CPU 720 co-ordinates to read and write data records and application instructions to and from storage 730.

[0042] CPU 720 also interfaces with communications module 760, which comprises network interface controller (NIC) 762. NIC 762 transmits and receives data to and from network 200.

[0043] Computer system 70 also features an interface 740 as indicated which operates to interface with a user. CPU 720 connects to various elements of the interface 740 via an input/output bus (not shown). Display processor 741 is connected to CPU 720, which is also connected to the display 742.

[0044] Keyboard 743 and track pad 744 (alternatively, mouse 744) also connect to CPU 720, which provide input directly from a user to CPU 720, in response to what is depicted on display 742.

[0045] Audio controller 745 is connected to CPU 720, which processes between auditory signals and audio data sent to and received from CPU 720. Audio controller 745 connects to speaker 746 and microphone 747. Audio controller 745 receives auditory output from the connected microphone 747, as well as generating auditory output to speaker 746.

[0046] Bluetooth module 748 is connected to CPU 720 and provides near range wireless communications with external devices (not shown).

[0047] Camera processor 749 is connected to CPU 720 and also camera 750 for receiving visual input that is processed to be sent to CPU 720.

[0048] Computer system 70 is configured to be externally powered, with power distributed for operation for each of the described components.

[0049] Tablet 80 is not specifically described as the hardware level, though is equivalent is very many respects to smartphone 90.

[0050] Smartphone 90 has an application processor 910. Application processor 910 is powered by power supply 920. Other elements of smartphone 90 described below are also powered internally by power supply 920. Power supply 920 has a power management module 922 which communicates with application processor 310 and interconnects with battery 924. Power supply 920 is periodically recharged by an external supply (not shown) as required-via wired connection or inductively.

[0051] Storage 930 comprises volatile memory 932 and non-volatile memory 934, each of interoperate under direction from the application processor 910. The application processor 910 reads and writes data from and to volatile memory 332 and non-volatile memory 934.

[0052] Communications 940 comprises a RF (radio frequency) transceiver 942 which interoperates with application processor 910 through baseband processor 944 to co-ordinate transmission and reception of RF signals, principally establishing connection to a mobile network, such as via 4G or 5G protocols, for native voice communications and internet communication. Baseband processor 944 receives signals from the RF transceiver 942 and provides application processor 910 with corresponding data, and conversely receives data from application processor 910 and generals a corresponding signal for RF transceiver 942.

[0053] Other communications modules connected to application processor 910 include WLAN module 946, Bluetooth module 948, GPS module 952 and NFC module 954.

[0054] Interface 960 of the device 900 comprises a display processor 961 connected to the application processor 910, which is in turn connected to the display 962. Touch screen controller 963 is associated with display 962 and connected to application processor 910. Touch screen controller 963 receives user input, typically prompted by the content show to a user on the display 962.

[0055] Audio controller 964 connected to the application processor 910 processes between auditory signals and audio data sent to and received from application processor 910. Audio controller 964 receives auditory output from the connected microphone 965, as well as generating auditory output to speaker 966.

[0056] Camera processor 967 is also connected to application processor 910 and camera 968 for receiving visual input that is processed to be sent to application processor 910.

[0057] Physical buttons 969 such as on/off, volume or display brightness control provide input to application processor 910. Application processor 910 also provides signals to a haptic engine 971 that provides tactile feedback or kinaesthetic communication to a user 10 holding smartphone 90.

[0058] Smartphone 90 embodies varies sensors 980 beyond those noted as part of the interface 960. These include accelerator 982, gyroscope 984, proximity sensor 986, barometer 988, ambient light sensor 992, and LiDAR scanner 994.

[0059] FIG 4 depicts in schematic form a logical architecture corresponding with the physical architecture of FIG 1.

[0060] Backend logic module 120 comprises as depicted a pre-qualification checks module 122, an income validation module 124, an expenses module 126, an assets module 127 and liabilities module 128.

[0061] The backend logic module 120 draws on data sources that are in the preferred but non-limiting embodiment described as external to the platform server 100. These data sources comprise external data sources 202. Other data sources supplying the backend logic module 120 or receiving data via the backend logic module 120 comprise the content of the database 130.

[0062] Database 130 stores a property loan application record comprising in part user-reported property loan application data 140 sourced through the web app 20, and originating with the user 10 via their device 70-90. Property loan application records further comprise data sourced from servers 210-260 as described above, referred to generally as external data 150. External data 150 may have been, processed, supplemented or otherwise manipulated at the backend logic module 120, in particular by modules 122-128 described above, drawing on data from servers 210-260 as also described above.

[0063] One data source supplying backend logic module 120 is expenses profiles 155 stored at database 130. Expenses profiles 155 store indicative data for household cost of living expenses for different locations and demographic profiles, periodically updated as required. Another data source supplying the backend logic module 120 is locality profiles 160 stored at database 130. Locality profiles 160 store indicative information against which loan applications cannot proceed, or other information that may inform predetermined logic rules, as described in further detail below.

[0064] The external data services 202 supply data as follows. Banking data service 262 supplies banking data to the income validation module 124, the expenses module 126 and the liabilities module 128. Property valuation service 212 supplies property valuation data to pre-qualification checks module 122, and also to assets module 127. Credit reporting service 232 supplies credit reporting data to liabilities module 128.

[0065] Financial identification service 252 supplies financial identification data to a decision engine service 242, which is in communication with the backend logic module 120. Decision engine service 242 is external to the server 100 in a preferred embodiment. Decision engine service 242 reaches a conclusion of approval result 1 or decline result 0 in connection with a property loan application.

[0066] The database 130 depicted in FIG 1 comprises property loan application data records as described.

[0067] Property loan application data records accumulate data as a property loan application progresses and comprise user self-reported data 140 as well as external data 150.

[0068] User self-reported data 140 comprises data indicated in TABLE 1 directly below.

TABLE 1

" Personal identification data for loan applicant (name, date of birth, address, email, telephone, driver's licence no, for example) • Loan property details (address or postcode, estimated value, for example)

• Transaction banking login credentials for loan applicant (username, password, for example) * Living expense data for loan applicant estimated by loan applicant (monthly or weekly, categorised, for example)

[0069] External data 150 comprises amongst other data an automated property valuation model report received from a property valuation server 210 for the loan property, a cost of living report extracted from the cost of living profiles 155, an automated credit report received from the credit report server 230 for the loan applicant, an automated banking transaction report received from the banking data server 260.

[0070] External data 150 comprises data as indicated in TABLE 2 directly below.

TABLE2

• Automated property valuation model report for loan property from property valuation server 210 (estimated value and forecast standard deviation, for example)

• Automated credit report for loan applicant from credit report server 230 (credit score, existing and closed liabilities, repayment history, public court records, for example)

• Automated banking transaction report from banking data server 260 (line item transactions that may be optionally categorised)

[0071] FIG 5 is a flowchart of steps involved in operation of the automated decision engine described and depicted in connection with its preferred embodiment. User 10 opens web app 20 in step 402, using a device 70-90. Web app 20 presents to user 10 a choice between alternative scenarios of whether the user 10 is seeking a property loan to buy a property or refinance a property in step 404. The user 10 indicates their response, and in either instance identifies the property in question in step 406.

[0072] A locality check is also performed in step 408. Locality check 408 references the property location (for example, street address, or simply postcode), and compares to locality profiles 160 depicted in FIG 4.

[0073] Locality profiles 160 maintains an index of postcodes that are deemed high risk and as a consequence will lead to a decline result 0.

[0074] An automated valuation model (AVM) check is also made in step 408. This involves the platform server 100 making a request to property valuation server 210 for an automated property valuation report comprising at least the estimated value and standard deviation of the estimated value. This may be respect of a specific address or alternatively, for a selected postcode. In this alternative, loan applicants can use a postcode location (for example) as a proxy for an actual address to secure pre-approval.

[0075] Should either the locality check or AVM check fail the applicable threshold requirement in step 408, the result is a decline result 0. Otherwise, processing continues to step 410, and user 10 enters the loan details for the property loan sought.

[0076] Then, a loan-to-value ratio (LVR) check is made in step 412. LVR is a ratio of the numerical amount of the loan amount sought to the value of the underlying property asset. A loan value ratio (LVR) table stored at database 130 indexes acceptable loan to value ratios, which reflect a ceiling ratio, which decreases as the forecast standard deviation increases.

[0077] Selected localities or postcodes in some embodiments may attract stricter than baseline requirements in lending criteria. This may manifest as selected localities shaping a lower ceiling limit on acceptable LVR ratio for particular localities during LVR check 412.

[0078] A check that is outside of predetermined criteria results in a decline result 0 at this point.

[0079] Otherwise, processing proceeds to step 414, in which a user 10 enters their relevant identifying details as prompted by the app 20. Certain of those details are authenticated in step 416, such as telephone number or email address by a simple second factor authentication process.

[0080] Following authentication, processing continues to step 418, in which a record of loan application details is created or accessed if already in existence. Details may be confirmed by user 10.

[0081] Then, a credit reporting service check is made in step 420, followed by a check of a cost of living profile in step 421, and a check of bank transaction data in step 422.

[0082] These steps 420-422 retrieve a credit report and also a cost of living profile and a banking transaction report on the user.

[0083] An income assessment in step 423 compares user-reported data regarding income and income data extracted from the automated banking transaction report.

[0084] An expenses assessment in step 424 compares data items from the cost of living profile, which benchmarks living expenses for people meeting the profile of a loan applicant, and user-reported data concerning living expenses, as well as cost of living data extracted from banking transaction data.

[0085] An assets assessment in step 425 may for example compare property assets indicated in user-reported data 140 with data from the automated property valuation report.

[0086] A liabilities assessment in step 426 compares existing and closed liabilities data from the automated credit report with user-reported liability data, and optionally also data extracted from the automated banking transaction report.

[0087] A serviceability assessment is made in step 428, drawing upon the preceding assessments in steps 423-426.

[0088] At this point, a decline result 0 may follow if predetermined logic rules regarding the income, expenses, assets and liabilities of the user suggest that the user will be unable to service the loan amount.

[0089] Otherwise, decision engine logic is applied in step 430, which may also give a decline result 0.

[0090] A financial identification check is made in step 432 which if successful gives an approval result 1, otherwise a decline result 0.

[0091] FIGS 6A-6C are collectively a sequence diagram that makes clearer the details of the process outlined in connection with the steps depicted and described in the flowchart of FIG 5.

[0092] A user 10 opens web app 20, and identifies the user 10 at the backend server 110 in step 502, with acknowledgment by reply in step 504. The user 10 indicates whether the loan sought is for a property that is being bought (or refinanced as the case may be) in step 506, which an acknowledgement returned in step 508.

[0093] The user enters the identifying details of the property in step 510, which is recorded at the backend server 110. The backend server 110 requests an automated property valuation report 140 for the identified property of the property valuation server 210 in step 512. The property valuation server 210 returns the automated property valuation report in step 514.

[0094] Backend server 110 requests a locality profile based upon property location supplied by user 10 in step 516 from database 130, and a locality profile is sent back from database 130 to backend server 110 in step 518.

[0095] The backend server 110 performs a locality and AVM check in step 520, as described in connection with FIG 5. As a consequence, the backend server 522 arrives at a decline result 0 (or OK), indicated at web app 20 in step 522. Otherwise processing continues, and further loan details are supplied from web app 20 in step 524. The backend server 110 makes a LVR check in step 526, as also described in connection with FIG 5.

[0096] A decline result 0 or OK results in step 528 as a consequence of the LVR check.

[0097] Further user details are supplied from the web app 20 to backend server 110 in step 530, and the user authenticated in step 532-such as by secondary channel such as mobile phone or email account.

[0098] A property loan application record is requested by the backend server 110 from database 130 in step 534, or created if not already in existence. The property loan application record is returned to backend server 110 in step 536.

[0099] The backend server 110 requests an automated credit report for the user 10 as loan applicant from a credit report server in step 538, and an automated credit report returned from the credit report server 230 to the backend server 110 in step 540.

[0100] The backend server 110 requests an expenses profile for the user 10 from the database 130, which the database 130 returns to the backend server 110 in step 544.

[0101] The backend server 110 then requests an automated banking transaction report (also known as a 'bank scrape') for the user 10 from the banking data server 260 in step 546. The banking data server 260 returns the automated banking transaction report to the backend server 110 in step 548.

[0102] The backend server 110 updates the property loan application record in step 550.

[0103] Backend server 110 performs an income assessment in step 552, expenses assessment in step 554, assets assessment in step 555, liabilities assessment in step 556 and serviceability assessment in step 558-all as described above in connection with FIG 5.

[0104] Serviceability assessment in step 558 may result in a decline result 0 being transmitted to web app 20 (not shown).

[0105] Otherwise, the backend server 110 requests a final decision of decision engine server 240 in step 560 using selected items from the property loan application record. Decision engine server 240 returns an 'OK' result in step 562. Alternatively, a 'not OK' result may be returned, which also results in a decline result 0 from backend server 110 to web app 20 (not shown).

[0106] Backend server 110 then requests confirmation of financial identity of financial identification server 250 in step 564. Financial identification server 250 returns an 'OK' result to backend server 110 in step 566, which leads to backend server 110 issuing an approval result 1 to web app 20 in step 568.

[0107] Should the financial identification server 250 instead return a 'not OK' result in step 566 indicating a failure to verify financial identity then backend server 110 instead gives a decline result 0 in step 568 (not shown).

[0108] The foregoing description and accompanying drawings concerns a preferred embodiment only. As will be appreciated many modifications or alternatives may be made without departing from the present invention.

Claims (5)

CLAIMS The claims defining the invention are as follows:

1. A computer-implemented method for providing financial services comprising: receiving user-reported property loan application data at a platform server, via a network from a web app accessed by a user using a personal computing device, and storing at the platform server the user-reported property loan application data as part of a property loan application record; transmitting from the platform server to one or more data servers external to the platform server at least one data request based upon one or more data items selected from the property loan application record, and receiving in response from the data server an automated data response at the platform server; storing the automated response data at the platform server as part of the property loan application record; and calculating at the platform server whether or not the property loan application represented in the property loan application record meets one of multiple threshold requirements for automated approval of the property loan application by applying predetermined logic rules recorded at the platform server to one or more data items selected from the property loan application data record.

2. A computer-implemented method as claimed in claim 1, further comprising: transmitting from the platform server a data request to a property valuation server in respect of a property location, and receiving in response from the property valuation server an automated property valuation report at the platform server, the automated property valuation report comprising at least an estimated value and a distribution metric; storing the automated property valuation report at the platform server as part of the property loan application record; and calculating whether or not the property loan application represented in the property loan application record meets a threshold requirement for automated approval of the property loan application by applying predetermined logic rules recorded at the platform server to a determination of loan to value ratio based upon the automated property valuation report and the loan amount recorded in the property loan application record.

3. A computer-implemented method as claimed in claim 2, further comprising: transmitting from the platform server a data request to a credit report server in respect of the user, and receiving in response from the credit report server an automated credit report for the user at the platform server, the automated credit report comprising at least a credit score of the user; storing the automated credit report at the platform server as part of the property loan application record; and calculating whether or not the property loan application represented in the property loan application record meets a threshold requirement for automated approval of the property loan application by applying predetermined logic rules recorded at the platform server to the credit score.

4. A computer-implemented method as claimed in claims 2 and 3, further comprising: transmitting from the platform server a data request to a banking data server in respect of the user, and receiving in response from the banking data server an automated banking transaction report for the user at the platform server, the automated banking transaction report comprising at least banking transactions of the user; storing the automated banking transaction report at the platform server as part of the property loan application record; calculating whether or not the property loan application represented in the property loan application record meets a threshold requirement for automated approval of the property loan application by applying predetermined logic rules recorded at the platform server to the property loan application record as regards income, expenses, and liabilities of the user.

5. A computer-implemented method as claimed in claim 4, further comprising transmitting from the platform server a data request to a decision engine server in respect of selected data items of the property loan application record, and receiving in response from the decision engine server an automated decision as to compliance with a credit policy stored at the platform server.

PROPERTY VALUATION CREDIT REPORT DECISION ENGINE SERVER 210 SERVER 230 SERVER 240

FINANCIAL IDENTIFICATION SERVER 250

​SMARTPHONE 90 BANKING DATA NETWORK 200 SERVER 260

BANK SERVERS 270 ​TABLET 80 ​BACKEND SERVER 110

BACKEND LOGIC MODULE 120

​COMPUTER 70 ​

​APP 20

​DATABASE 130 APPLICANT 10 ​PLATFORM ​ SERVER 100

DISPLAY TRACKPAD SPEAKER MICROPHONE CAMERA 742 744 746 747 750

​ DISPLAY KEYBOARD AUDIO CAMERA BLUETOOTH 748 PROCESSOR 741 743 CONTROLLER 746 PROCESSOR 749

INTERFACE 740

NETWORK INTERFACE CENTRAL PROCESSING VOLATILE ​ CONTROLLER 762 UNIT 720 MEMORY 732

COMMUNICATIONS 760 NON-VOLATILE MEMORY 732 ​ STORAGE 730

NETWORK 200

DISPLAY SPEAKER MICROPHONE CAMERA HAPTIC ENGINE 962 966 965 968 971

INTERFACE 940 ​ DISPLAY TOUCH SCREEN AUDIO CAMERA PHYSICAL PROCESSOR 961 CONTROLLER 963 CONTROLLER 946 PROCESSOR 967 BUTTONS 969

ACCELEROMETER APPLICATION 982 BASEBAND PROCESSOR 910 WLAN 946 ​ PROCESSOR 944 ​ GYROSCOPE 984 ​ BLUETOOTH 948 ​ PROXIMITY ​ SENSOR 986 RF TRANSCEIVER POWER VOLATILE ​ GPS 952 MANAGEMENT 922 MEMORY 932 942 BAROMETER 988 ​ BATTERY NON-VOLATILE NFC 954 924 MEMORY 934 AMBIENT LIGHT 992 COMMUNCIATIONS 940 POWER SUPPLY 920 STORAGE 930 LIDAR SCANNER 994

SENSORS 980

DATABASE 130

EXPENSES ​USER SELF-REPORTED PROFILES 155 DATA 140

LOCALITY PROFILES ​EXTERNAL DATA 160 150

APPROVAL 1

​ DECLINE 0 PRE-QUALIFICATION CHECKS MODULE 122

BANKING DATA SERVICE 262 INCOME VALIDATION MODULE 124

PROPERTY VALUATION SERVICE 212 EXPENSES MODULE ​ 126 DECISION ENGINE SERVICE 242 CREDIT REPORTING ASSETS MODULE SERVICE 232 127

FINANCIAL IDENTIFICATION LIABILITIES MODULE SERVICE 252 128

​EXTERNAL DATA SERVICES 202 BACKEND LOGIC MODULE 120

​USER OPENS APP 402 SERVICEABILITY CHECK CRS 420 ASSESSMENT 428 404

BUY OR REFI? CHECK DECISION ENGINE COL DATA 421 LOGIC 430

IDENTIFY IDENTIFY PROPERTY 406 PROPERTY 406 CHECK BANK FINANCIAL IDENTITY TRANSACTION DATA 422 CHECK 432

LOCALITY CHECK + INCOME AVM 408 ASSESSMENT 423 APPROVAL 1 DECLINE 0

EXPENSES DECLINE 0 ENTER LOAN DETAILS 410 ASSESSMENT 424

ASSETS LVR CHECK 412 ASSESSMENT 425

ENTER USER LIABILITIES DETAILS 414 ASSESSMENT 426

AUTHENTICATION 416​ CREATE OR ACCESS LOAN DETAILS 418

​

APP ​BACKEND ​DATABASE ​PVS ​CRS ​DES FIS BDS 20 110 130 210 230 240 250 260

USER ID 502 OK 504

BUY ​506 OK 508

PROP ID 510

REQ PROP LOCATION 512 AVM REPORT 514

REQ LOCALITY 516 LOCALITY PROFILE 518

LOCALITY + AVM CHECK 520

DECLINE OR OK 522 LOAN ​ DETAILS 524

LVR CHECK

APP ​BACKEND ​DATABASE ​PVS ​CRS ​DES FIS BDS 20 110 130 210 230 240 250 260

DECLINE OR OK 528 USER DETAILS 530

AUTHENTICATE 532 REQ APPLN 534

APPLN RECORD 536 REQ CREDIT REPORT 538

CREDIT REPORT 540

REQ EXPENSES PROFILE 542

EXPENSES PROFILE 544

REQ BANK SCRAPE 546

BANK TRANSACTION REPORT 548

UPDATE APPLN RECORD 550

APP ​BACKEND ​DATABASE ​PVS ​CRS ​DES FIS BDS 20 110 130 210 230 240 250 260

INCOME ​ ASSESSMENT 552 EXPENSES ​ ASSESSMENT 554 ASSETS ​ ASSESSMENT 555 LIABILITIES ​ ASSESSMENT 556

SERVICEABILITY ​ ASSESSMENT 558 REQUEST DECISION ENGINE 560 ​

DECISION ENGINE OK 562

REQUEST FINANCIAL ID VERIFICATION 564 ​ FINANCIAL ID OK 566

APPROVAL 568