# Project export: DIAL(*)

This document was generated by HackStack to give an AI agent context about a hackathon project. Sections are labeled with their provenance; content marked as truncated was cut to keep this document small.

## Project metadata

- Hackathon: Cal Hacks 12.0
- Tagline: How we booked an SF Luxury Hotel for $158
- Devpost: https://devpost.com/software/dial-4rqzc3
- GitHub: not linked
- Video: https://www.youtube.com/embed/32DSdO1V-z0?enablejsapi=1&hl=en_US&rel=0&start=&version=3&wmode=transparent
- Result: winner (Rox: Best Use of Rox; Conversion: Best Use of Conversion; Crater: Play-Do Prize)
- Team: contributor stats unavailable

## Devpost submission (written by the team)

### Overview

More on SQL Optimizations

### Inspiration

We were stranded in San Francisco on Saturday night without housing, and decided to let a fleet of AI agents find us a room. What started as a quick experiment turned into a full-stack system for autonomous negotiation, data-driven decision making, and natural communication.

### What it does

DIAL (*) is an optimized customer service agent designed for data-heavy use cases. It makes phone calls, speaks naturally, queries live databases, and learns from every conversation to improve future performance.

### How we built it

Customer Service Agent Functionality Phone Call Capability – Twilio Custom Voice + State Machine – ElevenLabs API Gateway – ngrok (for Express/TypeScript backend) Backend Logic – Express server managing call flow and SQL queries SQL Optimizations Created specialized query plans tailored for repetitive agent tasks Cached common lookups and indexed query-heavy tables for >100× faster response times

### Challenges we ran into

Building a robust state machine that avoided hallucinations while staying conversational was hard. Integrating ElevenLabs’ voice with live Twilio calls also required solving real-time latency issues.

### Accomplishments we're proud of

Our agent called over 100 hotels and got real human responses from 40+ of them. It negotiated room prices down to $158 + tax — and landed us a bed for the night. Our SQL optimizations cut decision latency from seconds to milliseconds.

### What we learned

We learned how to merge LLM-style intelligence with deterministic systems. We also saw firsthand how structured databases and state machines can ground AI behavior, creating a system that’s both smart and reliable.

### What's next

for DIAL (*) We plan to scale DIAL into a general-purpose customer service infrastructure — allowing businesses to plug in their own data and voice models for seamless, human-like service automation.

## README (from the GitHub repository)

No README available.

## Detected evidence (automated analysis)

No repository was indexed for this project. Claimed technologies below could not be checked against code.
- Express (technology) — claimed on Devpost, not found in the code
- Node.js (technology) — claimed on Devpost, not found in the code
- SQL (language) — claimed on Devpost, not found in the code

## Codebase structure

No repository index available.

## Key source files

No repository index available; no source files included.