Night Quota – a Halloween game about my hometown Monsters work one night a year. In Kharkiv, that night ends at 11 p.m. Not because of monsters. Because of the curfew.

Pick one of four monsters, walk the streets, scare people, knock on doors, and hit your quota before dawn. After eleven the streets empty and the job gets harder.

I’m from Kharkiv and I miss it, so I rebuilt it as a game. 18 districts, each started from my own photo of a real street and drawn as a constructivist poster, the style the city is known for. Every district has a local with a story about Kharkiv. The passers-by are people I love: my friends, my daughter as a little witch, and my cat Lisa, who hides in every district.

Features used: Rive CLI + GPU Canvas (both), Scripting, Data Binding, State Machines.

How it’s made • Rive CLI + Claude Code: I’m a designer, not a game developer. I described the game, the agent wrote it, I watched the live preview and corrected it. • RML for the scenes, Luau scripting for the logic: the night clock, the districts, the scares, the dialogues. • GPU Canvas: one WGSL shader over the whole game. Curfew frame, searchlights, air-raid flash, a camera shutter for cat photos, the dawn report developing out of a halftone raster. • Data Binding drives the HUD, State Machines run the buttons. • Rive Editor to play it and check every screen. • Higgsfield for character art, from my own references and photos.

Play it on desktop, with a keyboard and mouse. Give it about 20 seconds to load, the whole city is in there.

The combination of clawed hands and the friendly priest is awesome! 😂🔥

Your own secure AI workforce, built on your own data.

Most SMEs pay for one expensive, generalist AI that knows nothing about their business. We're flipping that: a small, focused AI assistant for each department (HR, Accounts, Marketing, Sales, and Operations), each trained on your own business data.

How we build it:

The result is lower cost, higher accuracy, and an AI that actually understands how your business runs.

We're piloting this internally at Indigloo first. If you run an SME and want to see a department assistant in action, DM me. 🚀

🔗 indigloo.co

For the internal pilot, I'd keep a held-out question set per department and compare RAG alone with RAG plus fine-tuning. Which department do you expect to benefit most from tuning, and what errors would you want it to reduce?