Coffee, Code, Collab – Build together, ship by the end of the day.

Bring a project idea or start fresh alongside other builders turning prompts into production-ready systems. No slides, no panels — just laptops, caffeine, and people who ship.

You’ll:

• Work shoulder-to-shoulder with other AI engineers and founders.
• Stand up a live backend that reacts to real data and events.
• Share, swap, and extend ideas in the open hacking session — everyone leaves with something running.

What We’ll build

Pick up the next layer of Claude Code: building real backends, not just prompts.

By the end of the morning, you’ll understand how to make AI coding assistants ship working systems with memory, triggers, and APIs — and when to trust, extend, or override them. You’ll also explore how Claude Code can orchestrate full-stack workflows — and how to take control when you need to.

The best part? Hanging out with other people who build for real.
Swap notes on your stack, trade shortcuts, and see what others are experimenting with while you hack on your own idea. Everyone’s building something different — that’s the point.

Schedule

• 10:00 AM Check-in & Coffee
• 10:30 AM – 12:30 PM Guided build → first event → memory → endpoint
• 12:30 PM – 1:30 PM Catered Lunch & Networking w/ fellow builders
• 1:30 PM – 4:00 PM Open Hacking & Collaboration (choose a track, extend, add data sources)
• 4:00 PM – 5:00 PM Project Demos & Wrap-Up

Who This Is For

• Software engineers, ML / agent builders, and indie hackers with a project idea or repo.
• New to Raindrop is fine; comfortable with a terminal and running code is expected.

Not for: no-code tools, slide decks, or pitching. This is hands on.

What You’ll Use

Raindrop gives you a declarative way to define event-driven backends (sources, transforms, storage, triggers) so you can scaffold services fast — without boilerplate.

Think: on X → do Y → persist to Z → expose at /api/.
source “github_webhook”
bucket “events”

on “push” from “github_webhook” do
write to “events”
post to “/api/notify” with { repo, author, files_changed }
end

You’ll use it to build something useful — fast.

FAQ

Do I need to know LiquidMetal or Raindrop?

• No. You’ll learn by building.

Will this lock me in?

• No vendor lock-in here — you’ll leave with patterns you can reuse anywhere and a repo that runs locally.

What do I need installed?

• Git + Python or Node. We’ll share a one-liner setup to get you ready fast.

Sponsors

LiquidMetal AI is building Raindrop, the first AI-native framework designed for AI coding assistants to build production-grade backend systems.

Through the Raindrop MCP (Model Context Protocol), AI coding assistants like Claude Code and Cursor learn the entire development playbook—from architecture and PRD generation, through test-driven development cycles, to deployment and validation. The framework provides declarative infrastructure, AI-native primitives

(SmartMemory, SmartBuckets, SmartSQL), and complete versioning so AI assistants can ship working applications with persistent memory, multimodal storage, and natural language database access—all while following best practices and maintaining full observability. Raindrop turns AI coding assistants into production-ready development partners that build real systems, not just prototypes.