We gave Team SoloMid the product its technology had been waiting for, four AI creation studios and an in-game overlay, designed before a line of code existed.
Team SoloMid is one of the best-known organizations in competitive gaming. Andy Dinh founded it in 2009, and by 2022 Forbes ranked it the most valuable esports company in the world at $540 million, ahead of every other org on the list.
Its product arm matters as much as its teams. TSM owns Blitz, a desktop companion that coaches competitive players in real time across League of Legends, Valorant, Teamfight Tactics, CS2, Fortnite, and Marvel Rivals, reported at more than ten million active users in September 2024 across eight countries. After the organization stepped back from competitive esports in 2023, that software business became its engine.
Vibe was the next bet. In February 2024, four months before this engagement began, the company advertised publicly for an AI Research Engineer to lead "custom model development for a game modding platform," at a base salary above $500,000. The conviction, the budget, and the underlying technology were already in place. The product was not.
Working technology, and nothing a person could actually use.
Engineering could already reskin a character into Batman inside Grand Theft Auto V. What did not exist was any way for a human being to ask for it. There was no interface, no flow, no precedent to copy, and no settled definition of what the product was.
There was no competitor flow to benchmark, no established pattern for prompting a 3D asset into a running game, and no user expectation to meet. Every screen had to be argued from first principles and then defended.
Several senior stakeholders each held a different picture of the product, and direction reversed between sessions: make it feel like Fortnite, make it simpler, make it like nothing on the market, make it like everything on the market. The product was designed four separate ways before it settled.
You cannot swap a human character for a dragon without breaking the game that has to animate it. The original ask, generate any 3D model and drop it into any game, had to be pared back in some directions and expanded in others to stay buildable.
Team SoloMid tried repeatedly to hire designers in-house and could not retain them. There was no design capacity, no product practice, and no brand to work from. The visual system had to be built from typography and components up.
Modders today leave the game, visit a specialist site, download a file, install it, and hope. Making that accessible to a child without insulting the people who already do it was the core design tension of the entire project.
Drag the handle to compare the workflow modders live with today, a specialist site, a download, an install and a risk, against the Vibe overlay, opened over the running game with a keyboard shortcut.
With no precedent to design against, the research had to build one. WANDR scored the entire adjacent market, recruited from a community that is genuinely hard to reach, and wrote four user journeys that gave the team something concrete to argue about.
Meshy, Leonardo, Tripo, Spline, Luma Labs, Kaedim, ArtStation, 3DFY, Masterpiece X and others, scored across 29 criteria covering generation, customization, file formats, marketplace, revenue sharing, community, onboarding, and pricing. Almost everyone could generate an asset. Almost nobody solved discovery, sharing, and in-game use together. That gap was the opportunity.
The first wave returned seventeen responses and not one respondent belonged to a modding community. Later waves brought fraudulent applicants and no-shows, flagged and removed one by one. Recruitment was rebuilt and re-run four times before the panel was defensible.
A 14-year-old Minecraft player, a 22-year-old creative professional, a 32-year-old 3D artist, and an independent game studio, each mapped across seven stages with the emotion, pain point, opportunity and outcome named at every step. The spread from schoolchild to studio is what forced the product into separate studios rather than one generic tool.
High-school student who plays Minecraft with friends and customizes skins. Wants to make something cool without advanced skills, and quietly worries the AI will not understand what he types.
Creative professional across Fortnite, GTA V, and Minecraft who wants one signature style expressed in all of them. Found the platform through a social ad. Sharing is not an afterthought, it is the point.
Professional 3D artist who already mods. Skeptical that AI can meet his standards and wary of losing creative control, but drawn by the time it saves and the marketplace he could sell into.
Independent game studio weighing the platform as production tooling. Wants to import its own models, cut asset-creation cost, and iterate fast, and needs the whole team onboarded, not one person.
Clean, intuitive interfaces where every step is streamlined to cut friction. The measure was whether a ten-year-old could start modding without reading anything.
The same visual and functional logic on mobile and desktop, so generating an asset on the move and applying it at the desk feel like one product rather than two.
Open with the magic. Put AI-generated work in front of people immediately, before signup, before explanation, because seeing what is possible is the only reliable way to explain a category that does not exist yet.
Discoverability and sharing built into the structure, not bolted on. Users rate, remix, and build on each other's creations, which is how the asset library grows faster than any one team could fill it.
The deepest experience belongs inside the game. Overlay editing tools let people change what they are looking at while they are looking at it.
Every version of the brief described the problem as difficulty: modding is hard, so make the tools easier. The research pointed somewhere else. Modding is not primarily hard, it is separate. You stop playing, open a browser, find a specialist site, download a file, install it, and hope nothing breaks. That break in context is what keeps the hobby at a few million skilled people instead of everyone who owns a game.
So the product's center of gravity moved from the studio to the overlay: a panel opened over the running game where you point at a character, describe what you want, and watch it change. Once that became the spine, the rest of the platform had a reason to exist and an order to be built in.
Team SoloMid ended the engagement with a coherent product where it had an abstract ambition, and a design record complete enough to hand to engineers and to investors.
This engagement delivered a design concept rather than a shipped product, so no post-launch performance metric exists. Every figure above is drawn from WANDR's project record: the signed statement of work, the competitive benchmark, the recruitment tracker, the journey maps, the delivered Figma files, and the closure letter of November 20, 2024.
Working with Wandr was a game-changer for our product vision. They quickly grasped the complexities of creating an AI-powered platform in an emerging market. Their expertise in user research and iterative design, paired with their ability to adapt to a fast-paced startup environment, resulted in an intuitive and user-friendly design that exceeded our expectations.
