PROJECT OVERVIEW

Kontxt

Kontxt is an AI-powered platform for preventing mobile messaging fraud, developed by RealNetworks.

  • RealNetworks is an enterprise company providing artificial intelligence and computer vision–based products. Founded in 1994, it is based in Seattle, Washington, United States.

Kontxt - Home

Home

PROBLEM DEFINITION

How to Stop Fraud and Spam Messages

According to statistics, the average person receives around 50 text messages per day, but not every notification is something they actually want.

Fraud attempts, spam, and smishing have become increasingly common, often appearing as simple texts, links to call or message a number, or URLs leading to suspicious websites or app downloads. These unwanted messages create frustration, waste time, and can lead to serious financial or security risks.

At the same time, important messages, especially those related to safety, urgency, or expectations are often missed or never delivered.

How can we detect and prevent harmful messages while improving the overall messaging experience?

DESIGN PROCESS

Design Process

Since Kontxt was a new product with no existing user data, I began by identifying user types, understanding their needs, and uncovering the challenges they faced when interacting with messages, based on informed assumptions I later validated through usability testing.

User Research

  • Personas

  • Journey map

  • Competitive analysis

  • Usability testing

Design

  • HMW questions

  • User flows

  • Wireframes

  • High-fidelity design/prototype

  • Final design

  • Design handoff

USER RESEARCH

Persona

Kontxt serves two groups of users: Customers (mobile network operators, aggregators, businesses) who use it to filter messages, and Subscribers who receive those messages.

This phase focused on Subscribers: people whose experience is shaped entirely by Kontxt’s decisions, even though they never interact with it directly.

To keep the team aligned on who we were designing for, I synthesized the research into a primary persona: their goals, frustrations, and mental models.

Journey Map

Building on the previously established persona, I created a journey map to illustrate how subscribers experience both wanted and unwanted text messages. This helped me uncover pain points, explore potential solutions, and inform design decisions.

Competitive Analysis

Competitors with machine-learning capabilities offer many of the features explored in this project, but most lack a comprehensive yet simple web-based solution.

DESIGN

How Might We

I created How Might We questions to align on both the customer’s goals and the subscribers’ needs, guiding design decisions that cater to all types of users.

  1. How might we improve the messaging experience by stopping fraud and spam content for subscribers?

  2. How might we provide a platform that would analyze and classify messages while allowing customers (mobile networks operators, aggregators, businesses) to access all critical information?

  3. How might we enable customers to strengthen loyalty and grow revenue?

Wireframes

During wireframing, I explored information architecture options that matched how customers expected messages to be organized and accessed.

Pictured below are my initial explorations of the Home/Dashboard page.

User Flow

After aligning on the proposal with the team, I created a user flow that helped us define the functionality and the information architecture.

USER RESEARCH

Usability Testing

I created a prototype of the proposed solution and led in-person usability testing with 4 participants, one from each future partner company:

  • Vodafone (mobile network operator)

  • Syniverse (aggregator)

  • Cloudli (business communications provider)

  • Iconectiv (telecom infrastructure)

All participants successfully completed key tasks, such as monitoring messages and understanding analytics.

SOLUTION

Platform for Preventing Spam and Fraud Messages

An AI platform that helps improve mobile content deliverability and detects spam/fraud over SMS, voice and IP channels. It helps mobile network operators, aggregators, businesses deliver a better messaging experience for their subscribers.

Using machine learning, the technology can analyze and classify most message types: Two Factor Authentication, Customer Support, Promotion, Emergency Alert, Notification, Grey Route, Fraud, Spam.

The end result is a spam-free messaging experience for subscribers and increased trust in mobile network operators and brands, delivered through a quote-based enterprise solution tailored to each client’s infrastructure.

How Kontxt Works

DESIGN

Final Design

A comprehensive web app with a clean and easy to use interface. White space throughout the app is used to balance design elements and convey grouping.

Message types are presented in various colors to indicate a message type, for instance, green is used for authentication while red for emergency. Having strong and contrasting colors help in better readability, sense of hierarchy and space.

Design System

I’ve designed a scalable design system, based on the atomic design methodology that covers all use cases that have been indentified for this project.

Usability Heuristics

  • Visibility of System Status

  • User Control and Freedom

  • Consistency

  • Recognition Rather than Recall

  • Flexibility and Efficiency of Us

Challenges and Trade-Offs

Message Classification

The challenge was how to visually distinguish each message type using a unique accent color to signal both type and urgency. While usability testing validated our color choices, the trade-off is that users may interpret colors differently, so the distinction might not be universally intuitive to users.

Below is a preview of the design system’s responsive web components, alongside screens from the web app.

Design System

Kontxt - Home

Home

Kontxt - Analytics

Analytics

Kontxt - Messages

Messages

Kontxt - Support Requests

Support Requests

Kontxt - Classification

Classification

Kontxt - Notifications

Notifications

IMPACT

KPIs

Product analytics indicate strong adoption and engagement among customers. Customer reports also show a significant reduction in spam and fraudulent messages received by subscribers.

82%

spam and fraud reduction rate

81%

adoption rate

90%

engagement rate

How We Measured Success

  • Spam and fraud reduction rate (82%):Percentage reduction in verified spam and fraud messages, comparing network logs before and after the launch of the Kontxt filtering system. Kontxt analytics provided message-level data used to identify and validate threats over time.

  • Adoption rate (81%):Percentage of users who interacted with the app after gaining access, out of the total user base (post-launch).

  • Engagement rate (90%):Percentage of active users performing meaningful actions in the app monthly, such as monitoring messages and interacting with analytics dashboards.

All usage-based metrics were tracked via Google Analytics.

PROJECT DETAILS

Project Details

  • Role: Senior Product Designer, UX Researcher

  • Company: RealNetworks

  • Industries: AI, Telecom

  • Market: B2B, SaaS

  • Team: Senior Product Designer (me), Product Manager, Engineering

  • Framework: Scrum

  • Tools:

    • Design: Figma, Miro

    • Project management: Jira, Confluence

    • Product analytics: Google Analytics

  • Year: 2018

  • Platform: Web

  • Link: Kontxt.com