Published in


Na Lie — A SiriusLabs Case Study

What is Na Lie?

Na Lie is the latest way to verify the “truthfulness” of information from the internet using a sophisticated machine learning library, it helps you verify information. From job vacancies to social media posts, Na Lie analyses provided phone numbers, text, and email addresses, and confirms their level of authenticity.

It is an Android app that verifies and filters fake messages. It is a real-time validation system that properly verifies text messages while ensuring data privacy. The aim is to proactively detect and prevent text-based financial fraud and fake messages from successfully soliciting information and funds from unsuspecting users.

The Challenge

Information flows freely in today’s technological advanced era. Several sources provide data and reports to the public and most times, it is difficult to tell which one is true or fabricated. In Nigeria, there is a lot of information in circulation on major social media platforms, be it news reports, job vacancies, snapshots of information etc, and due to the accessibility to mobile devices, people can view these easily. The major issue is knowing what information is real or fake.

What We Did

The client, an MTN and Data Science Nigeria associate, came to us with an existing machine learning model, and a need to build this solution. During the design and development phases, we trained the model to process different data sets accurately. We created minimalistic wireframes and went through three iterations for the initial designs. After aligning on the low-fidelity designs, we shared the high-fidelity prototypes with the client, and finalized the Android app development. The aim was to build a simple and functional product for users so they focus on the authenticity of information they intend to confirm or verify.

Research & Process

After discussions with the clients, we understood the process the machine learning library follows to collect data in order to confirm and verify it. Since most users view information through text or image formats, we figured they should be able to upload or copy and paste the information in the app for the AI model to process.

We used the existing ML model and trained it using lots of existing scams and legitimate information. Once a user wants to verify information, we run this through the model to get a classification for the data.

Code snippet from Machine Learning library

User Journey and Flowchart

We mapped out a possible user journey for basic interactions within the app. The goal was to put ourselves in the user’s shoes (or thumbs) and understand their thought process. We jotted the journey down on paper and designed the flowchart as shown.

User Journey
User Flowchart


Next step was to create basic wireframes for the app UI. We had to take the basic features into consideration so that the design could make sense before proceeding with iterations.

Initial wireframes

Design Iterations

We went through a couple of design iterations, mostly with the layout and colour palette. After reviews, we translated our wireframes to mid fidelity screens.

Mid-fidelity screens

The clients decided we go with a green colour variant, so we went ahead with a few adjustments and implemented their preferred design.

Colour Palette & Typography

The colour palette was chosen based on the clients’ request and certain components in the app, such as the rating scale. We figured it should have a colour range; from green, down to orange and red, to indicate the authenticity of the information provided.

We also chose Google Sans as the app typeface to give it a modern and sleek appearance.

Colour Palette

Final Designs

We made use of Clak!, an illustration pack from LSGraphics to display some states in the app. The final designs of the key screens are shown below.

Profile Screen
Contribution Screen
State for added contribution under review
Successful Contribution


With the ease of accessibility to information, it is necessary to know how to filter through the bulk and find the truth. The Na Lie app helped us understand the urgency in verifying and contributing the truth and helping the global knowledge pool grow. Designing it to be as straightforward as possible will possibly engage users to add more input for the Machine Learning model to learn and filter information more accurately.

If you have an android device, check out the app on playstore! Thanks for reading and following!




Musings, thoughts and process of how we achieve things at SiriusLabs

Recommended from Medium

Brief Explanation on Corner Pooling

Figure 1

Let an AI generate your next lecture or speech for you

How to Configure a Facial Recognition System for Face ID Authentication

When do we need Machine Learning?

A Review of Object Detection Models

Lasso/Ridge and Compressing Meaning

Week #1 — Identification of Artists by Their Paintings with Machine Learning

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Aderinsola Akintilo

Aderinsola Akintilo

Multidisciplinary Designer — Striving to make Nigeria a better country

More from Medium

A research on the product listing and product description pages of different types of e-commerce…

Treiner — Connecting Football Organizations to Coaches

Prose & Poetry

UX Case study: Card sorting for 3Shape Community