Introduction
Little Red Riding Hood is a third-person action-adventure game built in UEFN, inspired by the classic fairy tale. The game is split into three levels, each with its own atmosphere and gameplay mechanic that evolves as the story progresses. Every level has its own cutscenes and voice acting to bring the story to life.
My Roles:
Little Red Riding Hood is a solo project. I handled the full experience end-to-end, with collaborators contributing voice performance only.
Level Design: designed three standalone levels with distinct mechanics, atmospheres, and difficulty curves, supported by a level design document and flow map for each chapter
Quest Design: built and sequenced all objectives, triggers, and progression systems in UEFN
Narrative Design: adapted the source material into interactive beats, redesigned key story moments for gameplay, and directed cutscene structure and dialogue
System Implementation: implemented all gameplay systems including dialogue management, item mechanics, AI companion, countdown timer, and cinematic toolset
Level Design
Three Levels, Three Mechanical Identities
Each level needed its own mechanic that matched the emotional tone of that story beat — rather than applying a uniform gameplay loop across the whole experience.
Level 1 — The Farmlands & Mill Bright, open, and clearly signposted. Cutscene-driven and deliberately relaxed — an onboarding level focused on establishing the world before any tension sets in.
Level 2 — The Forest A closed loop set at night. Players collect items to unlock a new path, but the final key item is a carriable object — holding it prevents attacking and slows movement. A jump on the return route blocks the safe inbound path, forcing players through new terrain while encumbered. Rather than locking players into stealth, multiple completion options are available: stay stealthy, drop the item to fight, then pick it back up. Playtesting showed the stealth-only approach felt restrictive, so opening up the options kept tension without frustration.
Level 3 — Inside the Wolf A decaying pocket dimension under a countdown timer. Players fight their way out alongside grandmother as a companion, culminating in a boss fight — the sharpest escalation in the game.
Quest Design
Sequencing Objectives Across Three Chapters
Each level needed to feel like a self-contained chapter with a clear arc — without losing the player or letting pacing drift.
Before building, I wrote a level design document and flow map for each chapter, mapping every encounter, beat, and narrative moment in sequence. This made implementation straightforward and prevented structural problems that are painful to fix mid-build.
In UEFN, objectives are sequenced through the task tracker, trigger volumes, item granter, and dialogue managers. Playtesting after each level informed iterative adjustments — trimming sections that ran too long, patching out-of-bounds areas, and rebalancing difficulty spikes. A central lobby with portals connects all three levels, keeping each modular and independently testable.
Narrative Design
Adapting a Fairy Tale for Interactive Play
The challenge was identifying which story beats work interactively and which needed to be rebuilt from the ground up.
The original structure is preserved — departure, encounter, consumption, escape — but the implementation of each beat was redesigned. The wolf becomes Mr. Lupin, a character the player meets through direct dialogue. The ending was the biggest departure: the source material's resolution offers no player agency. Instead, Red and grandmother fight their way out together, turning a passive rescue into an earned escape.
Each level opens and closes with cutscenes and voiced dialogue, delivering story context at natural pause points without competing with gameplay.
Technical Design
Managing Systems Across Three Levels
The core challenge was coordinating multiple managers in parallel levels, for example, ensuring objective completion reliably triggered the next dialogue, world reaction, or unlock at the right moment.
Patrol AI in Level 2 uses manually set paths to control sightlines and pressure in the stealth section. Level 3 runs on UEFN's native bodyguard AI for grandmother, paired with a countdown timer and environmental damage volumes.
Cutscene implementation came down to camera control — locking player input during cinematic volumes and timing camera track transitions to prevent characters from wandering into frame or moving erratically during playback.