HackStack

Project Info

Plannr

Devpost

This project did not submit a demo video on Devpost.

Inspiration

As hackathon organizers ourselves, we know firsthand how overwhelming and chaotic planning can get. Finding sponsors, booking venues, managing transportation, designing websites, ordering swag, handling outreach, and running logistics, all while keeping everything on schedule, requires hundreds of emails, calls, and spreadsheet updates. Inspired by the theme of this hackathon, we asked: Why not build an AI system that can do most of this for us? That’s how Plannr was born: an intelligent AI-powered assistant that automates the hardest parts of organizing hackathons.

What it does

Plannr is an AI Hackathon Organizer Assistant powered by a system of specialized agents that automate and streamline every stage of hackathon planning. It helps you: Find and reach out to sponsors, write contextual emails, follow up, and log progress Search for venues, call vendors with Vapi, negotiate pricing, and track details Coordinate bussing and catering, automatically call around for quotes and log everything Generate a content calendar with posts, reels, and emails using Claude and Gemini Schedule and log Instagram posts, with AI-generated visuals Manage outreach for mentors, speakers, and communities Auto-generate websites, hackathon applications, and answer FAQs Everything is coordinated through a central spreadsheet system and connected via n8n workflows that use agents from Claude, Gemini, Fetch.ai, Unify, Orkes, Groq, Vapi, and letta.ai.

How we built it

Claude (via Fetch.ai uAgents): Content generation, email writing, route suggestions, sponsor research Gemini: Generated visuals (posts, reels), swag design, website copy Vapi: Outbound voice calls for bus quotes, venue calls, meal vendor outreach n8n: Our orchestration layer; we built a suite of agent-based workflows across outreach, logistics, and marketing Google Sheets API: All metadata is logged to editable and synced spreadsheets Groq, Orkes: Integrated to power backend services, real-time decisioning, and scalable processing Frontend: Simple interface to add and edit hackathon metadata (in progress)

Challenges we ran into

Workflow explosion: Managing over 10 parallel n8n workflows with asynchronous agents was messy Voice API hiccups: Setting up Vapi assistants and linking IDs dynamically in n8n required a lot of debugging LLM agent coordination: Syncing Claude with Fetch.ai agents and letting Gemini handle visual generation within time constraints Rate limits and APIs: Many services had undocumented rate limits or unstable behavior Time constraints: Trying to ship a multi-agent production system in under 24 hours with only two people.

Accomplishments we're proud of

Built a modular, multi-agent automation system for planning events Designed workflows that can run independently or in sequence Used voice agents to actually call real vendors (bus, venue, catering) Auto-generated a full week of Instagram content (posts and reels) Created our own Claude sponsor research and outreach agent Integrated multiple AI platforms together in one cohesive system

What we learned

How to orchestrate multi-agent workflows using n8n, Claude, and Fetch.ai How to integrate voice automation using Vapi and parse responses How to structure event metadata centrally for agents to read and write from That AI can reduce hours of planning into minutes with the right system design That even the best agents need solid UX, traceability, and logging

What's next

Let student hackathon teams onboard and use our planner assistant One-click snack, merch, and bus ordering with pre-filled AI quotes Improve location-based planning with map integration Deploy a chat interface so users can ask for tasks like "find me 3 venues" Offer integrations with Devpost, Luma, Vercel, and Slack Create a public library of reusable AI workflows for event organizers

Analysis

Compare with all teams

View

Metric

Figures cover GitHub contributors during the hackathon window. A co-authored commit counts in full for each author, so per-member totals add up to more than the whole-team figures.

Technology

Found in codeClaimed only
  • CSSIn code
  • HTMLIn code
  • JavaScriptIn code
  • ReactIn code
  • SupabaseIn code
  • Tailwind CSSIn code
  • TypeScriptIn code
  • Google GeminiClaimed

7 of 8 appear in the indexed code. 1 claimed on Devpost could not be matched to code, which may simply mean the tool leaves no trace in the repository.

AI coding agents

No AI coding agent signals were found in this repository.

Detected from committed agent config files and commit authorship. Absence of a signal is not proof an agent was unused.

Codebase size

Source size

304 KB

Source files

71

Counts recognized source files only; vendored directories, binaries and lockfiles are excluded, so this is smaller than the repository on disk.

0 stars