ARGOS Score

Introduction

The ARGOS Score is a score calculated based on the user's eKYC submission data. If the result meets or exceeds the customer-set threshold, the submission shall be automatically approved. By leveraging this functionality, businesses can replace traditional manual processes with a faster and more efficient automated system.

But worry not! Submissions below the threshold still receive manual assistance. After all, ARGOS is always focused on reducing customer hassles.

Background

The eKYC process using ID cards involves various variables. Especially in the recognition process using computer vision, there are many instances where OCR doesn't accurately extract information due to factors such as light reflection, damage, and camera quality.

Currently, many users experience inconvenience, such as having to start over the process from the beginning due to inaccurate OCR extract.

It's evident that as the number of retry attempts for a user's eKYC approval increases, users tend to leave the service!

The ARGOS Score was designed to prevent user attrition by comprehensively analyzing submitted images and text information to produce a reliability score, thereby supporting accurate and swift user onboarding.

ARGOS Score Threshold

The ARGOS Score Threshold that can be set in the dashboard serves as a criterion for determining the level of automation in the KYC (Know Your Customer) process. In other words, it sets the standard for deciding whether to automatically approve a submission or to proceed with a manual review.

If the set score is high, more manual reviews will be conducted. On the other hand, if the set score is low, there will be more automatic approvals. For example, if you set the threshold at 70 points, submissions with an ARGOS Score exceeding 70 will be automatically approved, while those with scores of 70 or below will be subject to manual review, followed by either approval or rejection.

*If not set separately, the default threshold of 70 points will be applied.

Explanation with Examples

The ARGOS Score represents the 'reliability' of identity authentication, determined based on our own AI algorithm.

It calculates the probability of accuracy by consolidating factors such as recognition of the ID card and user verification. By checking the eKYC details submitted by the user on the ARGOS dashboard, you can see the precise ARGOS Score displayed. Here, we have a submission for eKYC using an ID card from “LAURIS KRISTIANS PUGEJS.”

• Name: LAURIS KRISTIANS PUGEJS

Name	Birthdate	Document Number	Personal Number	Gender	Nationality	Issue Date	Expiry Date
LAURIS KRISTIANS PUGEJS	1984-11-15	AMO903jd9f	DL3N390F0M	Male	USA	2020-03-10	2030-03-09

When proceeding with eKYC, digital information is extracted from the analog data of the identification card, including various OCR processes. Depending on the results of this digital information extraction, the ARGOS Score will vary.

While the final eKYC outcome may be the same, the ARGOS Score can change based on the results of the digital information extraction process.

The ARGOS Score is calculated by consolidating the following information:

1. Similarity of the recognized ID card to the Template.

2. Reliability of the values extracted by the OCR.

3. Reliability of the final values entered by the user (verified with truth verification, BAR CODE, MRZ, etc.).

4. Similarity between the final values entered by the user and the OCR results.

Below is an example table of the ARGOS Score:

Case 1. ARGOS Score: 99 points

1. ID Card Recognition: Successful

2. OCR Extraction & Reliability:

   Category	Name	Birthdate	Document Number	Personal Number	Gender	Nationality	Issue Date	Expiry Date
   OCR Extracted	LAURIS KRISTIANS PUGEJS	1984-11-15	AMO903jd9f	DL3N390F0M	Male	USA	2020-03-10	2030-03-09
   OCR Reliability	99	99	99	99	99	99	99	99
3. User Confirmation Trustworthiness

   1. Verification Successful for Truthfulness: Name, Date of Birth, Document Number, Personal Number, Date of Issue

   2. MRZ Verification Successful: Date of Birth, Document Number

   3. BAR CODE Verification: None applicable.

Case 2. ARGOS Score: 80 points

1. ID Card Recognition: Successful

2. OCR Extraction & Reliability:

   Category	Name	Birthdate	Document Number	Personal Number	Gender	Nationality	Issue Date	Expiry Date
   OCR Extracted	LAURIS KRISTIANS	1984-11-15	AM09O3jd9	DL3N390F0M	Male	-	2020-03-10	2030-03-09
   OCR Reliability	85	80	70	88	90	-	90	88
3. User Confirmation Trustworthiness

   1. Verification Successful for Truthfulness: None applicable.

   2. MRZ Verification Successful: Name, Date of Birth, Document Number.

   3. BAR CODE Verification: None applicable.

Case 3. ARGOS Score: 70 points

1. ID Card Recognition: Successful

2. OCR Extraction & Reliability:

Category	Name	Birthdate	Document Number	Personal Number	Gender	Nationality	Issue Date	Expiry Date
OCR Extracted	LAUMS KRISHAN	1984-11-15	AM09O3jd9	DL3N390F0M	-	USA	2020-03-10	-
OCR Reliability	60	50	-	99	-	75	75	-

1. User Confirmation Trustworthiness

   1. Verification Successful for Truthfulness: None applicable.

   2. MRZ Verification Successful: Name, Expiration Date.

   3. BAR CODE Verification: None applicable.

Case 4. ARGOS Score: 65 points

1. ID Card Recognition: Successful

2. OCR Extraction & Reliability:

Category	Name	Birthdate	Document Number	Personal Number	Gender	Nationality	Issue Date	Expiry Date
OCR Extracted	KRISTIANS PUGEJS	-	AMO903jd9f	DL3N390F0M	-	-	2020-03-10	2030-03-09
OCR Reliability	50	-	80	70	-	-	80	60

1. User Confirmation Trustworthiness

   1. Verification Successful for Truthfulness: None applicable.

   2. MRZ Verification Successful: None applicable.

   3. BAR CODE Verification: None applicable.

Case 5. ARGOS Score: 50 points

1. ID Card Recognition: Successful

2. OCR Extraction & Reliability:

Category	Name	Birthdate	Document Number	Personal Number	Gender	Nationality	Issue Date	Expiry Date
OCR Extracted	AURIS PU	-	AMO903jd9f	-	-	-	-	2030-03-09
OCR Reliability	70	-	90	-	-	-	-	90

1. User Confirmation Trustworthiness:

   1. Verification Successful for Truthfulness: None applicable.

   2. MRZ Verification Successful: Date of Birth.

   3. BAR CODE Verification: None applicable.

Case 6. ARGOS Score: 30 points

1. ID Card Recognition: Successful

2. OCR Extraction & Reliability:

Category	Name	Birthdate	Document Number	Personal Number	Gender	Nationality	Issue Date	Expiry Date
OCR Extracted	LAURIS KRISTIANS PUGEJS	-	-	-	-	-	-	-
OCR Reliability	90	-	-	-	-	-	-	-

1. User Confirmation Trustworthiness:

   1. Verification Successful for Truthfulness: Date of Birth.

   2. MRZ Verification Successful: None applicable.

   3. BAR CODE Verification: None applicable.

Case 7. ARGOS Score 10 points (No identification recognized)

1. ID Card Recognition: Successful

2. OCR Extraction & Reliability:

Category	Name	Birthdate	Document Number	Personal Number	Gender	Nationality	Issue Date	Expiry Date
OCR Extracted	-	-	-	-	-	-	-	-
OCR Reliability	-	-	-	-	-	-	-	-

1. User Confirmation Trustworthiness:

   1. Verification Successful for Truthfulness: None applicable.

   2. MRZ Verification Successful: Personal Number, Expiration Date.

   3. BAR CODE Verification: None applicable.

Best Practice

100% Automated Processing

If you wish to eliminate the waiting time for manual review and only proceed with real-time automated processing, you can consider using the 'Automated Processing Only' option in the dashboard along with adjusting the ARGOS Score threshold.

When the 'Automated Processing Only' option is enabled, all KYC applications that would otherwise fall into the Pending (manual review required) stage are automatically rejected. This allows the user to retry the KYC process. By appropriately lowering the ARGOS Score threshold at this stage, more KYC applications are likely to be approved.

For Faster Onboarding

If you want to reduce user waiting time and expedite the onboarding process, you can set the ARGOS Score threshold to a lower value. This will increase the ratio of automatic approvals, while the accuracy can be verified post-approval.