Beaver Quest
A 2D top-down puzzle game built in Unity by a 5-person team (codename Eden V). Players control a beaver rebuilding dams across a forest, pushing cubes onto pressure plates to open gates and collect wood. My formal role was Level Designer, but on a team this small I also led stakeholder discussions, drove the initial mechanic and theme design, and rode herd on milestone deliverables.
Built across a four-month production cycle from prototype through launch.
Designed three levels end-to-end
Built Levels 1, 4, and 5 in Unity: the onboarding tutorial level, the cube-reuse difficulty spike, and the final puzzle climax. Owned each from whitebox through art-passed final.
Led stakeholder discussions
Acted as the team's primary stakeholder-facing voice for design reviews. Translated feedback into actionable changes and brought decisions back to the team for execution.
Drove initial mechanic and theme design
Played a primary role in the early ideation phase that landed on the beaver-and-dams theme and the cube/pressure-plate core loop. Helped the team converge from a wide pitch to a single coherent direction.
Owned milestone deliverables
Oversaw turn-in of most milestone assignments, tracked teammate task completion against deadlines, and kept us on the rails through Pre-Production, Vertical Slice, Alpha, Beta, and Launch.
Cross-cultural collaboration
Worked with a team of teammates from different countries with different native languages. Built a communication style that prioritized clarity, repetition, and confirmation over speed.
Playtested and iterated
Sat through stakeholder reviews of my levels early and often. Used the feedback to escalate puzzle complexity without losing the relaxed, low-pressure tone the team wanted.
Each of my three levels had a specific job in the difficulty curve: teach, escalate, and culminate. Top-down captures from the final build below.
Level 1, Section 1 – Teaching the basics
An easy-to-follow linear path forces the player to interact with each tutorial prompt in order: move, pick up a cube, place a cube. No optional branches, no opportunity to skip the lesson.
Level 1, Section 2 – First puzzle, first dam
The second half introduces the wood collection loop and the dam-building payoff. Designed to be solvable in as few moves as possible so a brand-new player still feels smart.
Level 4 – Cube reuse, hard
Opens with three back-to-back puzzles that force the player to reuse the same cubes to progress. The mechanic has been simmering since Level 1, this is where it finally gets pushed to its limits.
Level 4, Finale – Triple pressure plate intro
The last section introduces the triple-pressure-plate mechanic in a controlled environment. Player gets one clean encounter with it here so Level 5 can lean on it without being unfair.
Level 5, Section 1 – The starter cube
The hardest puzzle in the game. Players who notice they need the very first cube from the level entrance save themselves a long backtrack later. Designed as an optional optimization, not a hard gate, to reward attention without punishing the rest.
Level 5, Finale – Double triple pressure plates
The climax. A triple-red and triple-blue pressure plate sequence that uses nearly every cube in the level. Built as the game's biggest "I figured it out" moment before the final dam.
- Stakeholder iteration kills ego. Reviewing my levels with our stakeholder every week meant I never got too attached to any single puzzle layout. Cuts and rebuilds became routine instead of painful. The levels are better because I stopped defending them.
- Cross-cultural communication is a real skill. Most of my team came from a different country and spoke a different native language. Once I started prioritizing clarity and confirmation over speed, the team got faster, not slower.
- Playtest earlier than you think. Near launch we found a critical bug in Level 1 that should have been caught weeks earlier. As the level's designer I owned that miss. The next two projects I worked on, I scheduled playtests into the sprint calendar by default.
- Difficulty curves are budgets, not slopes. Designing Levels 1, 4, and 5 forced me to think about how each mechanic gets introduced, escalated, and finally combined. Every new system needs at least one safe encounter before it carries real weight.
- Process beats heroics. Our biggest milestone misses came from skipped Perforce hygiene and inconsistent naming conventions. By the end we ran a daily 5-minute submit/pull check at the start of every lab. The fix was boring. It worked.
Beaver Quest was the first project where I owned stakeholder-facing communication. The weekly review cadence I learned on a 5-person team carried directly into stakeholder reviews on a 27-person team (Kila: Hourbound) and a 54-person team (HardDriverz) the following year.
- Stakeholder communication. Ran weekly design reviews with the project stakeholder. The workflow is the same as milestone reviews with a publisher: capture feedback, convert it into tasks the team can act on, close the loop the following week.
- Scope discipline. Three levels meant constant decisions about what to cut to keep quality high. That decision-making is the same one producers make at every scope review.
- Cross-cultural team management. Built a communication style for a team where most members spoke a different first language. Directly applicable at any studio with distributed or international teams.
- Owning the deliverable, not just its content. I designed the levels and made sure they were checked in, integrated, and visible in the build before every stakeholder review. That coordination layer is the producer's job on a larger team.
