All Case Studies

Huntr LVM

Huntr LVM was designed to stress-test agentic AI applications against targeted security threats in a competition format. Break AskNova. Win Rewards.

Palo Alto Networks Summer 2025 Lead Designer
Prompt Injections Backdoor Threats Gamification
Huntr LVM — AI threat hunting challenge

Overview

The LMV Challenge gamifies security testing through a competition format with leaderboards, reward pots, and timed competitions — departing from traditional bug bounty models.

Focused on a single hosted challenge with multiple independent objectives. Front-loaded competition format, mechanics, and rewards to entice participation. Used automated detectors to evaluate attack success and populate leaderboards instantly.

User Goals

Security Researcher

"I want to understand the challenge format so I can decide whether to participate."

Competitor

"I want to see all objectives clearly explained so I know what to attempt."

Motivated Participant

"I want to track my leaderboard position and potential payout so I stay engaged."

Top Competitor

"I want verification to ensure my rewards are protected and legitimate."

Information Architecture

Homepage

Hero banner explaining LMV program, competition model, and rewards with clear call-to-action

Challenge Overview

List of objectives (tool misuse, margin manipulation) with status indicators

Leaderboard

Real-time rankings with usernames, scores, and payout estimates

Rewards

Visual prize pot breakdown and rank-based distribution

Design Solutions

Homepage

Hero statement connecting to competition concept, "How LMV Works" 3-step illustration (Join → Attack → Earn), primary call-to-action

Challenge Page

Flippable cards displaying context on front, attack goals and success criteria on back. Submit attack button per objective.

Leaderboard

Real-time rankings with dynamic prize allocation. Multiple toggles: Shortest Prompt, First Blood, Difficulty.

Rewards

Visual prize pot representation, bonus vs. primary objective explanations, example scenarios for reward maximization

Challenges & Solutions

Duplicate Prevention

Detecting duplicate submissions could have diluted leaderboard fairness. Designed mechanics where duplicates yield no primary objective points but still count toward bonus categories like shortest prompt.

Competition Model Explanation

Users were familiar with bug bounty, not competition format. Created simple onboarding visuals and FAQ content to bridge the gap.

User Motivation

Keeping users engaged during long competitions. Implemented live leaderboards with visible prize impacts and "First Blood" recognitions for early achievers.

Design System

Typography

Bold, modern sans-serif for headlines emphasizing energy and competition. Neutral sans-serif body for technical clarity.

Color Palette

Deep navy for trust, electric green for hacking/gamification, amber/orange accents for rewards and urgency.

Components

Challenge objective cards, navigation tabs, progress indicators, sortable leaderboard table, reward meter with prize pot visualization.