Fraud Investigation

My Role in AWS:

Redesigning AWS Fraud Investigation Platform

Lead UX designer

  • Lead cross-functional alignment across complex problem spaces to define product strategy & set design vision.

  • Design for scale, autonomy, and impact. Work with internal & external customers to understand the challenges & bring their voice to the table

  • Recommend design solution, create end-to-end user experience & collaborate with engineers along with AI/ML scientists for bringing design to life. Bringing AI, compliance, and usability into a single user experience.

Platform: AWS Commerce and Fraud Management Platform

Team: Product, Engineering, AI/ML, Fraud Operations

Impact: +58% investigation efficiency | +80% Task efficiency | +Investigator trust

Amazon Web Services (AWS) is the world’s most comprehensive and broadly adopted cloud offering with millions of active customers and tens of thousands of partners globally. AWS Fraud Prevention team offers various fully managed solutions to monitor customer’s businesses globally, enabling them to identify potentially fraudulent activities and catch more online fraud faster.

AWS currently offers legacy fraud investigation tool which is used by external partners and internal investigation team to monitor fraud at all Amazon & AWS businesses. The tool is complex to use & difficult to understand. Due to complexity, it is not efficient in investigations and hard to maintain.

Business Goals

Design a modern, scalable, and user-centered investigation platform that:

  • Reduces investigation time.

  • Empowers non-technical users with self-service tools.

  • Consolidates workflows across the fraud lifecycle.

  • Aligns with AWS’s internal security, compliance, and branding standards.

Current State and Challenges

To understand the scope of work, I conducted cognitive walkthroughs and 1:1 interviews with current investigators. This research provided great insights about who users are, their daily activities & pain points in using current process.

AWS’s legacy fraud investigation process was:

  • Visibility of information

    Many users complained about not able to quickly read thorough & understand the content due to too many colors and tables. This increases investigation time significantly for new users.

  • Workflow Fragmentation

    Users had to switch between fraud applications during the investigation. More than 50% users asked to consolidate info into one system.

  • Visually outdated

    Users mentioned that the screens don’t look modern and feel old with many colors & icons. It gave impression that AWS is not paying attention to experience & forcing users to learn old school system.

  • Developer Bottlenecks

    The steps for updating business workflow are unclear & complicated. PMs couldn’t update workflows without engineering help. They have to raise tickets to AWS teams to make any changes to their system & reply on their availability to resolve it.

  • Low trust

    The inefficient and cumbersome product delivered low trust and engagement from both internal and external teams.

Design Strategy

We kicked off a multi-disciplinary design workshop with stakeholders from:

  • Product

  • Engineering

  • AI/ML

  • Business Ops

Persona

Persona

Key workshop outcomes:

  • Agreement on pain points & priorities

  • Definition of a role-based system

  • Alignment on MVP and long-term vision


Investigator

Program Manager

Needs & Frustrations

Needs faster access to actionable data, hates jumping between tools.

Needs ability to configure workflows without dev dependency.

We mapped their daily investigation flows and identified moments of friction and opportunity.

Journey Map

Navigation

Defining investigation journey for users in the system

Defined how information will be structured. Program manager will have access to entire architecture and Investigators will access part of it based on their role.

Design Solution

  • Self-Serve Configuration flow

    Empowered PMs to manage their workflows without ticketing devs.

  • Unified Task View

    Combined all investigation data in a single screen. Eliminated tool-switching.

  • Role-Based Dashboards

    Delivered context-aware dashboards tailored to each persona to provide investigation insights & bring attention to important notification/tasks

  • Modern UI

    Redesigned with cleaner typography, fewer colors, and improved visual hierarchy.

Landing Pg

Final Mocks

Self-Service Configuration : Step-by-step configuration wizard for PMs

Complete Configuration

Work Stream Dashboard

Investigation Dashboard : Detailed case view with linked evidence, notes, and AI/ML fraud insights

Investigation Dashboard

Self-serve Configuration

Investigate

Impact

Stakeholder Feedback

“This is the first time the platform actually feels like it’s built for us.” – Fraud Analyst

“Our PMs no longer need to file tickets for every change. This is a game-changer.” – Program Manager