E-KYC
E-KYC Indonesia from Alfabeta verifies who your customers really are in one automated flow — ID capture, OCR, face match and liveness — so onboarding is fast, compliant and fraud stays out.
Request a demo
Alfabeta E-KYC is an electronic customer verification system built for e-KYC Indonesia use cases, using computer vision to confirm a customer's identity with minimal human intervention. It runs the full digital identity verification pipeline end to end: it captures a government ID, reads the document with OCR, matches the ID portrait against a live selfie, and confirms the person is genuinely present through liveness detection. This replaces slow, in-person, paper-based checks with a remote flow a customer can complete on their own device.
Document handling is driven by OCR KTP capture — optical character recognition reads the Indonesian national ID (KTP) and other documents automatically, extracting each field into structured data. Because the data is machine-read rather than retyped by staff, the process eliminates typos and human error from manual entry, speeds up onboarding, and cuts operational cost. Face match and liveness then work together: face matching compares the captured selfie to the ID photo, while liveness detection distinguishes a real, present person from a printed photo, replayed video or mask to defeat presentation and spoofing attacks.
The service is aimed at regulated onboarding where accuracy and auditability matter — banking, fintech, e-wallets, telecom and other digital services. In the Indonesian context this supports customer due diligence obligations and aligns with the regulator's framework for digital financial innovation, including OJK POJK 12/2017; the platform's audit trail records every step and outcome so those checks can be evidenced during compliance review. A configurable decision threshold — commonly a 75% face match threshold — lets each institution tune how strict the pass/fail bar is to fit its own risk appetite.
Every verification returns a structured result with the evidence behind each decision, and the whole capability is delivered through clean APIs so it drops into an existing onboarding journey rather than replacing the front end. Alongside faster client onboarding, institutions gain an eco-friendly paperless process, fewer manual errors, better working-capital efficiency and lower operational cost — while keeping a defensible record of how each identity was verified.
ID document OCR
Reads the Indonesian KTP and other national IDs automatically, extracting every field into structured data without manual typing. This removes the typos and rekeying errors that come from manual entry and speeds up onboarding.
Face matching
Compares the portrait on the ID against a live selfie to confirm the document belongs to the person presenting it. A configurable match threshold, commonly set around 75%, lets each institution tune strictness to its own risk appetite.
Liveness detection
Distinguishes a real, present person from a photo, replayed video or mask to stop presentation and spoofing attacks. It ensures the selfie being matched is a live capture rather than a stolen image.
Identity decisioning
Combines OCR, face match and liveness into a single pass or fail result with the evidence behind each decision. Institutions receive a clear outcome plus the underlying signals, not just a score in isolation.
Audit trail
Records every step and outcome for compliance reviews and regulatory reporting, supporting customer due diligence obligations under frameworks such as OJK POJK 12/2017. The trail turns each verification into defensible, timestamped evidence.
API integration
Drops into existing onboarding journeys through clean APIs, so you keep your own front-end and workflow. Verification runs as a service behind your app rather than forcing a rebuild of the customer experience.
Frequently asked questions
Which documents can the OCR read, and does it support the Indonesian KTP?
Yes. The optical character recognition is built to read the Indonesian national ID (KTP) and other identity documents, extracting each field into structured data automatically. Because the data is machine-read rather than retyped by an operator, it removes the typos and rekeying mistakes that come with manual entry and shortens the onboarding process.
How does the system stop someone using a stolen photo or a printed ID?
Two layers work together. Face matching compares the live selfie against the portrait on the ID to confirm they are the same person, and liveness detection determines whether the selfie is a genuinely present, live person rather than a printed photo, a replayed video or a mask. Together they defend against the common presentation and spoofing attacks used in remote onboarding fraud.
What is the face match threshold and can we change it?
Face matching produces a similarity result that is compared against a configurable threshold — commonly set around 75%. Each institution can tune that bar to match its own risk appetite: a higher threshold reduces false accepts at the cost of more manual reviews, while a lower one accepts more automatically. The threshold is a policy decision the deploying organisation controls.
Does E-KYC help us meet Indonesian regulatory requirements?
It is designed for regulated onboarding and supports customer due diligence obligations, aligning with the OJK framework for digital financial innovation including POJK 12/2017. Every verification step and outcome is captured in an audit trail, giving you defensible, timestamped evidence to present during compliance reviews and regulatory reporting. Alfabeta provides the technology; the institution remains responsible for its own regulatory decisions.
How does it fit into our existing onboarding app?
The capability is delivered through clean APIs, so it plugs into your current onboarding journey instead of replacing it. You keep your own front-end and workflow, and call E-KYC as a service to capture the ID, run OCR, perform the face match and liveness check, and return a structured pass or fail result with its supporting evidence.