Mariah Fiby Clado

Overview

STS Studio is the studio my partner and I run. We build games, operate servers, and make the tools that sit around them, with a two-person core and no outside funding.

The bet was simple: most small studios build one thing, launch it, and hope. We wanted the opposite: products that feed each other. A game that gives the community somewhere to live, a launcher that brings players in from another platform, and internal tooling so two people can operate like a bigger team.

I run the studio day to day. That means the environment and technical build on the game side, the product side of the launcher, partnership relationships, marketing, and the operational scaffolding that keeps a two-person team from dropping things.

What We Build

The Hardest Build: STS Avenue RP

Roblox enforces a hard ceiling of roughly 10,000 triangles per mesh. The architectural models we started from were built for rendering, not real-time play, and were nowhere near that.

• The base city map came in at roughly 3 million triangles
• The amusement park was over 7,500 objects and several million triangles
• The library scene arrived at tens of millions of triangles across 3,000+ objects
• The shopping mall was a single fused mesh, massively over limit

The obvious fix, decimation, was the wrong one. Aggressive polygon reduction destroys exactly what makes a roleplay world worth spending time in: railings, trim, signage, the detail players notice up close. The real problem was never "make this smaller." It was "make this legal without making it worse."

Splitting Instead of Shrinking

Rather than reducing geometry, I split it: separating meshes by material first, then chunking along a spatial grid until every piece cleared the limit. The shopping mall alone became over 300 compliant parts; the library resolved into roughly 3,000 meshes, each sitting under a working cap I set below Roblox's ceiling for safety margin.

Scripting the Pipeline

Doing this by hand across thousands of objects was never realistic. I wrote Python inside Blender to batch the separation, cap each operation to a triangle budget so it would not stall, compute accurate per-vertex bounding boxes for correctly scaling rotated objects, and export with the exact axis and scale settings Roblox expects.

Preserving Detail Where It Counts

Where geometry did need thinning, I used planar dissolve and selective chunking instead of blanket decimation, removing redundant coplanar faces rather than sacrificing silhouettes. Every exported piece was individually named so it stayed selectable and editable in Studio instead of arriving as an anonymous blob.

Making It Behave In-Engine

Import was the halfway point. On the Studio side I anchored several thousand parts to stop the city physically collapsing on load, corrected rotation offsets across imported sets, and grouped assets under clean parents so the hierarchy stayed navigable.

Building the Systems Around It

The map is only half a world. I built working escalators driven by cycling CFrames, a stadium jumbotron running a live in-game camera feed, a day/night cycle with keyframed lighting moods, animated neon on the Ferris wheel, and a server-side avatar scale rule so proportions stayed consistent across the city.

Auditing What We Imported

Free marketplace assets are a genuine security risk on Roblox. I audited incoming models and found and removed backdoor scripts, including one using a job-ID check to stay invisible while the place was open in Studio.

Running the Studio

Partnerships

Nexus Launcher only works if there are communities on the other end of it. I built and manage the partner relationships that give the launcher somewhere to send players.

Community

Our Discord is where players, partners, and testers all land. I handle structure, moderation standards, onboarding, and the events that give people a reason to come back between updates, building on years of moderating live chat at scale.

Marketing

Short-form video does the acquisition work: character reels and game teasers cut for TikTok. The goal is to show the world rather than describe it, and route interest into the Discord where it converts into actual players.

Operations

Two people cannot afford ambiguity about who owns what. I keep roadmaps, SOPs, and launch plans documented, and lean on AI-assisted workflows to compress the work that would otherwise need a third person.

Results

• Three products across two platforms, one live and two in active development, run by a two-person core.
• A full playable Roblox city built inside the engine's hard limits without blanket decimation.
• A repeatable Blender-to-Roblox pipeline that now handles new models in a fraction of the original time.
• Three partnered Minecraft servers and a hosting partner secured and maintained.
• Internal CRM in build to consolidate community, analytics, and marketing into one system.
• Backdoor scripts caught and removed before they ever reached a live place.

Skills Demonstrated

Key Takeaway

Running a small studio is mostly deciding what not to do. The wins came from constraints: a platform limit that forced a smarter pipeline, a two-person team that forced real documentation, and no budget that forced partnerships instead of spend.

Visit STS Studio ↗

Overview

Nexus Launcher is a custom Minecraft launcher built to make accessing Minecraft communities easier across Windows and macOS.

The idea started from a simple challenge: many players experience friction before they can join a server, from manual setup and file management to compatibility issues between different platforms.

One of the biggest challenges was accessibility for macOS users. Minecraft Bedrock Edition has not had an official native release for macOS, making it difficult for Mac users to participate in Bedrock-based communities.

Nexus Launcher was created to provide a smoother and more accessible experience for players while building a central platform that connects users with different Minecraft communities.

The project has grown beyond just a launcher. Nexus Launcher is becoming an ecosystem supported by partnerships, including three partnered Minecraft servers and one hosting partner, helping create a stronger foundation for future community growth.

The Ecosystem

Nexus Launcher is designed to become a central hub connecting players, Minecraft communities, and server partners.

Currently, the platform has:

• 3 partnered Minecraft servers
• 1 hosting partner

These partnerships allow Nexus Launcher to support multiple communities while giving players easier access to different Minecraft experiences through one platform.

Instead of players searching for servers individually and going through separate setup processes, the goal is to create a unified experience where communities and players can connect more easily.

The Problem

Minecraft has a large and diverse player base, but accessing custom experiences can still be complicated.

Mac users face additional limitations because Bedrock Edition does not have an official macOS version, which creates challenges for players who want to join Bedrock-based communities.

Players also commonly experience issues such as:

• Downloading the wrong files
• Installing incorrect versions
• Configuring settings manually
• Troubleshooting compatibility problems
• Following multiple setup guides

The goal was to reduce these barriers and make joining Minecraft communities simpler.

The Goal

Nexus Launcher was designed to:

• Improve accessibility for Windows and macOS players
• Simplify the setup process
• Connect players with partnered servers
• Create a central platform for Minecraft communities
• Build a scalable ecosystem for future growth

The Interface

The launcher is built around a simple idea: everything a player needs before they press play should be one screen away. Editions, versions, servers, and account, without leaving the app.

Visit the Nexus Launcher site ↗

Overview

As a Sales Representative at CSO Management, I carried my own book of customers in a high-volume digital sales environment, owning each account from first contact through the close and the repeat purchase after it.

The role was never order-taking. It meant reading buying signals, qualifying who was actually worth my time, handling objections in the moment, and building enough trust that customers came back to me by name.

The Challenge

Revenue depended almost entirely on relationship strength and follow-up discipline. With a large active book, the challenge was keeping every conversation personal enough to convert while running dozens of them in parallel.

Key challenges included:

• Keeping sales conversations personalized at scale
• Qualifying and prioritizing the highest-intent buyers
• Driving repeat purchases, not one-time transactions
• Hitting revenue targets without burning the relationship

My Approach

Selling on Relationship, Not Script

I dropped the one-size-fits-all pitch and led with discovery instead. Tracking each customer’s interests, past purchases, and response patterns let me open with something relevant, which consistently outperformed generic outreach.

Running the Pipeline in CRM

I used CRM to segment my book, track every deal stage, and keep full context between touchpoints. Disciplined notes and follow-up timing meant no warm lead went cold because I forgot to circle back.

Iterating on What Converted

I watched which openers, offers, and objection responses actually moved deals forward, then doubled down on them. Reading what each buyer valued, whether price, speed, or attention, and matching the pitch to it lifted close rate over time.

Results

• Closed approximately $379K in attributed sales through direct customer engagement.
• Carried a personal book of accounts in a fast-paced digital sales environment.
• Built repeatable habits around discovery, objection handling, follow-up, and CRM hygiene.
• Ran high-volume conversations in parallel without losing personalization or dropping follow-ups.

Skills Demonstrated

Key Takeaway

Selling well came down to understanding, consistency, and trust, not pressure. The same principles carry straight into sales, account management, and operations roles where communication is what actually moves revenue.