Crew Cash

• 

  Crew Cash

Inspiration

Student organizations, roommate groups, and campus clubs often struggle to manage shared finances. Tracking expenses across multiple Venmo accounts, spreadsheets, and banking apps typically leads to missed payments, unclear budgets, and frustration. Furthermore, group treasurers lack the necessary tools to accurately forecast their runway or safely invest idle club funds.

What it does

CrewCash is a collaborative, AI-powered financial operating system designed to make group finances transparent and engaging. Key features include:

• Cinematic Voice Briefings: A "Play Daily Briefing" button uses ElevenLabs AI to provide a radio-quality rundown of the group's financial health, runway, and pending approvals.
• Intelligent Financial Coach: Powered by Google Gemini, the coach analyzes spending velocity and anomalies to offer actionable advice, like forecasting exactly when a club will run out of funds.
• Cryptographic Audit Trails: Uses Solana blockchain anchors to create an immutable, verifiable ledger so past expenses cannot be silently altered.
• Investment & Market Readiness: Features an InvestView and Market Ticker Tape to simulate or track treasury investments for long-term savings goals.
• Dual-Approval Safeguards: High-value expenses automatically trigger a pending state requiring two approvals from crew members to prevent unauthorized purchases.

How we built it

• Frontend: Next.js App Router, React, Tailwind CSS, and shadcn/ui.
• Backend & Auth: Supabase (Postgres, Row-Level Security, Auth).
• AI & Voice: Google Gemini SDK and ElevenLabs TTS API.
• Blockchain: Solana web3.js for ledger anchoring.
• Visuals: Recharts for dynamic data visualization.

Challenges we ran into

One of our biggest hurdles was figuring out how to anchor financial data on the Solana blockchain without slowing down the user experience. We also spent a late night wrangling Supabase's Row-Level Security (RLS) to properly enforce our dual-approval system so that unauthorized users couldn't bypass the expense threshold rules. Finally, getting the ElevenLabs TTS to sound natural while reading dynamic, on-the-fly financial data from Gemini required a lot of prompt tweaking.

Accomplishments that we're proud of

We're incredibly proud of getting the "Play Daily Briefing" feature to work seamlessly—hearing an AI voice give a radio-quality rundown of our dashboard data was a huge "wow" moment for our team. We're also proud of successfully implementing the Gemini Coach for real-time anomaly detection and managing the technical complexity of integrating Solana for on-chain audit trails entirely within the hackathon time limit.

What we learned

We learned a ton about prompt engineering and the Gemini SDK while building the intelligent financial coach. We also got a crash course in Web3 by working with solana/web3.js for the first time, and we walked away with a much deeper understanding of how to structure secure Postgres databases using Supabase RLS.

What's next for Crew Cash

• Direct bank integrations via Plaid.
• Automated reimbursement payouts.
• Native mobile applications for iOS and Android.

Built With

• blockchain
• elevnlabs
• gemini
• python
• recharts
• sonola
• supabase
• tailwind
• tts-api.com