Saolix SaolixGroup
Saolix Technologies · Compliance Solutions

Saolix K2M. Know who you are onboarding, and keep the record.

Known2Me is eKYC and identity verification built as one system: consent, capture, checks, human review, appeals and audit. It is not a chain of separate vendors stitched together.

Run the K2M Simulator Download the designed report format Book a walkthrough (opens in a new tab)

Pre-release The simulator and sample report show how K2M is designed to work, with fictional applicants; parts are not yet in the current release. The screenshots below are from K2M itself.

01 · Claim

Verify the person, not just the document, and keep the record that proves it.

02 · Story

Identity checks are usually assembled from parts: an OCR tool here, a liveness vendor there, a spreadsheet for the audit. The gaps between them are where fraud gets in, genuine customers drop out and compliance fails. K2M handles the whole journey in one place. It asks for consent before any biometric step, captures the document and the person, and checks the identity against earlier applications. Decisions go to human reviewers, with a second reviewer required for high-risk sessions; evidence custody, admin actions and session history are written to audit tables protected by database triggers, and a rejected session can be appealed.

03 · Chain

Automated checks inform. Reviewers decide.

Automated checks score each capture against minimum thresholds. Reviewers make the decision, and high-risk sessions need a second reviewer.

The K2M decision chain: consent and capture, automated checks, the determinism boundary, a reviewer decision with four-eyes for high-risk sessions, and a recorded, appealable decision K2M · DECISION CHAIN Captureconsent · document · selfie Checks assistliveness · duplicates Decisionreviewers · four-eyes Recordedaudit trail · appeal CHECKS ASSIST · PROBABILISTIC PEOPLE DECIDE DETERMINISM BOUNDARY
04 · Capabilities

What K2M does today.

01

Consent-first capture

A guided flow that records consent before any biometric step, then captures the document and a selfie. The mobile app adds the back of the document and active liveness challenges.

intro · consent · document · selfie
02

Duplicate identity detection

Compares new applications against earlier ones and flags the same document image appearing across different sessions, for a reviewer to confirm or dismiss.

cross-session · reviewer resolves
03

Human review with four-eyes

Reviewers approve or reject from one dashboard. For sessions at a high assurance tier, or with a high risk score recorded in the review, any decision needs a second reviewer, and one reviewer cannot give both approvals.

dual approval · policy-based
04

Audit trail and appeals

Evidence custody, admin actions and session step history are written to audit tables where database triggers block edits and deletions. Staff can open an appeal against a rejected session; the reviewer who recorded the rejection cannot decide the appeal, and the outcome is recorded with reasoning.

protected audit · appeals
05 · Deployment

Self-hosted, no third-party KYC vendor.

K2M is built as a container stack you operate.

K2M system overview, pre-release: applicants use the web capture or mobile app, reviewers use the admin console, and your onboarding systems connect through the API. Inside your walls, K2M records consent before any biometric step, captures the document, selfie and liveness, and runs automated checks for document reading, face comparison and duplicates. Checks inform and people decide: reviewers make the decision, with four-eyes for high-risk sessions; a rejected session can be appealed to a different reviewer; and the audit trail blocks edits and deletions. K2M keeps its own database and evidence store, and document reading and face comparison run locally with open-source components. Registry checks, SMS one-time passcodes and sanctions lists use only the services you configure. A connection to Saolix Eagle is planned. No third-party KYC vendor is needed. K2M · SYSTEM OVERVIEW K2M service you configure planned PRE-RELEASE YOUR PEOPLE & SYSTEMS OUTSIDE K2M INSIDE YOUR WALLS · SELF-HOSTED Applicants web capture mobile app Reviewers admin console decisions · appeals Your systems onboarding API K2M Consent Capture Automated checks Reviewer decision Appeals Protected audit trail recorded before any biometric step document · selfie · liveness document reading · face comparison · duplicates four-eyes for high-risk sessions decided by a different reviewer edits and deletions blocked CHECKS INFORM · PEOPLE DECIDE Database sessions · decisions · audit Evidence store documents · selfies Local processing open-source components SMS passcodes one-time codes Official registries identity checks Sanctions lists UN · OFAC Saolix Eagle planned · wider screening Third-party KYC vendor not needed only the services you configure

Simplified overview of K2M's self-hosted design. K2M is in pre-release. Dashed boxes are outside services you configure; the dotted box is planned.

Self-hostedA container stack with its own database, cache and evidence store, run on infrastructure you control.
Local processingDocument reading and face comparison run inside K2M, using open-source components, without a third-party KYC vendor. Checks against official registries and SMS one-time passcodes use the external services you configure.
ScreeningWhen a session is sent for review, K2M can check the applicant against the UN and OFAC sanctions lists (where those list sources are configured) and writes the results to its application log. Wider PEP and adverse-media screening, and a connection to Saolix Eagle, are planned.
06 · Evidence

Don't take our word for it. Inspect it.

Start with any of these.

Try it · intended design

K2M Simulator

Follow a fictional applicant through K2M's intended design, from consent to decision: the threshold checks, the four-eyes review and the appeal.

Open the simulator ↗ Download · PDF · intended design

Designed report format

The Identity Verification Report K2M is designed to produce, for a fictional applicant: document checks, threshold checks, four-eyes review, appeal and audit timeline. The current release does not yet generate it.

Download the sample → On request

Live walkthrough

We walk you through K2M itself with a synthetic applicant, while you watch.

Book a session ↗

Screenshots from K2M itself

Taken from K2M running on our own machines, with synthetic applicants and a specimen ID card. Images are cropped for size.

K2M admin console: KYC review dashboard listing approved, rejected and in-progress verification sessions
Review dashboard. Every verification session with its status, policy, completed steps and decision.
K2M duplicate identity alerts: the same document image detected across several sessions, awaiting reviewer action
Duplicate identity alerts. A real detection: we reused one specimen card across sessions and K2M flagged each reuse.
K2M appeal detail: a rejected applicant's appeal pending review, with uphold or overturn options
Appeals. An appeal against a rejection, awaiting review; the outcome, uphold or overturn, is recorded with reasoning.
K2M capture flow consent step
Consent first. The capture flow asks for consent before any biometric step. The heading names the Kenya Data Protection Act 2019, which the step was designed around; it is not a statement of certified compliance, which depends on how each deployment is operated.
From Insights

Further reading.

Identity

Identity decisions people can challenge

Hard minimums, two reviewers and a fair appeal in K2M.

Read → AI governance

Why AI advises and rules decide

The three tests a regulated decision must pass.

Read → Architecture

The determinism boundary in practice

How Eagle lets AI help an analyst without ever letting it decide.

Read →
Contact · Saolix GroupPeople. Products. Possibilities.

Your next chapter.
Let’s build it together.

Explore a product, discuss a project, or find the right expertise. Tell us where you want to go. We’ll help you take the next step.

Send an enquiry Book a working session

Prefer email? info@saolix.com

Enterprise softwarePrivate AIEngineering & consulting