

Snowflake X Beta Fund X Evermind Agent & Token Economy Hackathon
โ๐ Every agent needs memory. Every agent consumes tokens. Build the infrastructure and economics that make intelligence scalable.
โToken Economy in the AI Era
โEvery AI product runs on tokens. Every token has a price. But scalable agents need more than inferenceโthey need persistent memory, context, and the ability to improve over time.
โBuilding a demo is easy. Running an agent in production without going broke or forgetting the user is the actual work. Tokens are the P&L: compute is the market, models are the suppliers, and the enterprise buyer is asking one question: is this intelligence worth what it costs?
โSnowflake sits at the intersection of enterprise data and AI execution: the worldโs data cloud, Cortex Agents, and inference next to the data companies already use.
โEverMind provides the memory infrastructure through EverOS, its open-source memory layer for AI agents. It enables agents to remember users, retain context across interactions, and become more useful over time.
โThis Week, every team is required to build with EverMind infrastructure. Projects should use EverOS as part of the agentโs memory, personalization, context, or learning layer, while leveraging Snowflake to build, operate, or analyze the token economy behind the product.
โBuild agents that remember. Build intelligence that is economically sustainable. Build systems that can actually scale.
โ๐ ๏ธ Join our Discord for credits, team formation, and support: Join Discord Server 1-2 Builders per team
โ๐ Schedule
โ9:00 AM Check-in & Breakfast
โ10:00 โ 11:00 AM Opening Remarks and workshops from Snowflake, Beta Fund, Evermind, voice cursor and UpScaleX
โ11:00 AM Build
โ12:00 PM Lunch
โ1:00 PM Build
โ4:00 PM Hard submission deadline
โ4:00 โ 5:00 PM Demos (3 min limit)
โ5:00 PM ๐ Audience voting and award ceremony
โFocus Tracks
โ๐ฐ Track 1 ยท Cost of Intelligence Build anything that makes AI cheaper to run. Prompt caching, model routing, batch inference, context compression, spend dashboards. Show a real % cost reduction in the demo.
โ๐๏ธ Track 2 ยท Value of Intelligence Build an AI product with real willingness-to-pay. Consumer app, prosumer tool, or B2B feature. Show one paying user (or a clear path to $10 ARR per user).
โ๐ Wildcard (Open Category) Anything else in the AI-token economy: agent budgeting, inference amortization, compute marketplaces, chargeback, model brokerage. Bring conviction.
โ๐ Prizes and Bounties
โ1st $600 | 2nd $500 | 3rd $400
โ๐๏ธ 1-hour 1:1 brainstorm with UpScaleX + $200 cash for the standout team.
โ๐ค Featured Partners & Developer Perks
โSnowflake โ the world's data cloud. Perk: Cortex Agents access + Snowflake credits per attendee + engineering support from the Snowflake solutions team.
โEverMind builds EverOS, the open-source memory layer for AI agents โ the piece that lets an agent remember who it's working with and get better the more it's used. It's a natural fit for education, where learning compounds over time and the best support grows with the learner. Find our team on-site all day with EverOS credits, a quickstart, and engineers ready to help you build.
โUpScaleX is a San Francisco & Palo Alto-based seed fund and foundersโ community investing in AI-native applications, agentic economies, and vertical AI, built to empower the next generation of AI innovators.