gStrEats EyAI

Inspiration

In the Indian state of Tamil Nadu, there are over 300,000 registered street food vendors who currently operate in a digital blindspot. gStrEats EyAI aims to digitally empower these street vendors and patrons alike, shedding light on these invisible micro-economies. Our mission is to provide vendors with the tools to broadcast their real-time availability and analytics to have causal insights, develop strategies, and improve their business operations, while helping patrons discover legendary food spots through data-driven insights rather than just word-of-mouth. gStrEats EyAI respects the centuries-old wisdom of street vendors while augmenting it with space-age intelligence.

What It Does

The Problem Space

Street food is the lifeblood of urban culture, yet its data remains "invisible", scattered across fragmented YouTube vlogs and ephemeral social media posts.

The Solution Stack

1. For Vendors: Operational Empowerment

• Digital Presence: Register shops and menus, and manage "Sold Out" status in real-time.
• Live Signaling: Broadcast high-precision location signals, providing a digital identity even without a formal address.

2. For Patrons: Discovery & Interaction

• Multilingual Voice Orders: Place orders seamlessly in English or Tamil.
• Food Genealogy: Trace the "ancestry" of a dish, explaining how a recipe traveled across borders to reach a specific street corner.
• AI Vision (Lens): Upload food images to receive instant history, calorie estimates (with disclaimer), and personalized shop recommendations.
• Crowd Intelligence: Access real-time community reviews and Predicted Footfall to estimate wait times.

3. For Vendors & Patrons: Single-Shop Analytics

• Success Reasoning: Visualize a shop's "Power Score" through factors like Location Gravity, Flavor Moat, Social Resonance, and Economic Fit.
• Safety & Logistics: Real-time data on Safety Indices (nearest police stations, crime index, and lighting) and Urban Logistics (bus stops, pedestrian pathways, and parking).
• Traffic & Synergy: Real-time traffic patterns and a Synergy Matrix that correlates weather data with business performance.

4. For Vendors & Patrons: Sector-Wide Intelligence

• Dynamic Node Discovery: "Wide-Band Scrapes" of the web and YouTube to find trending stalls, automatically converting addresses into precise map coordinates.
• Sector Analytics: Visualize area-specific flavor variance, legendary dish clusters, economic tiers, and customer segmentation.

5. For Developers: From Black Box to Glass Box

• Neural Operations Dashboard: A transparent view into agentic AI workflows.
• 3D Mesh Visualization: Watch specialized agents collaborate in real-time via a WebGL/Three.js interface.
• Scaling & Cost Transparency: Monitor agentic logs for infrastructure scaling and track real-time AI costs per transaction.

How we built it

Using Google AI Studio

• Frontend: React 19, TypeScript, Tailwind CSS
• Intelligence: @google/genai (Gemini 3 Pro/Flash Preview, Gemini 2.5 Flash Native Audio)
• Mapping: Leaflet.js with customized Voyager Dark/Inverted styling.
• Visuals: Three.js (3D Neural Coordination Mesh), Chart.js (Spatial Analytics).

Challenges we ran into

1. Rendering a high-quality 3D Neural Mesh (Three.js) that pulses in sync with AI state changes without draining browser resources.
2. Transforming raw data points into causal insights. Used Gemini 3’s Reasoning to build the Synergy Matrix and Success Reasoning Polar Chart.

Accomplishments that we're proud of

1. Live Signal Ownership: Vendors without formal addresses can now broadcast high-precision GPS signals, turning their physical presence into a verified digital business node.
2. Precision Discovery: Patrons discover legendary food spots through data-driven insights rather than just hearsay.
3. Agentic Orchestration: Successfully built a central Supervisor (Gemini 3 Flash). Developed a 3d(Three.js) visualization of the agent coordination mesh, allowing the developer community to observe the AI's internal thought process and task handoffs.
4. State Integrity: Developed a centralized state mesh that allows disparate agents to share context without direct communication, ensuring system-wide data consistency. ## What we learned Using AI Studio allowed us to focus on the "soul" of our agents—their personality, cultural grounding, and reasoning capabilities, while the platform handled the heavy lifting of API calls and management. AI studio not only helped in vibe coding, but also in testing the entire app for stability. Turned street knowledge into an app.

What's next for gStrEats EyAI

1. Virtual Queue Entry: Voice orders auto-enroll patrons with unique queue IDs and precise wait times. When the vendor marks the order "Ready", the patron's phone gets an instant notification. Eliminates physical crowding, boosts vendor throughput in busy areas.
2. Multimodal Video Ingestion (Vendor & Patron): Enable vendors to upload real-time video clips of their prep area to generate "Freshness Badges" via the Lens Agent. Lens Agent to generate real-time "Freshness Badges," directly solving the trust gap that often prevents new customers from trying street food. Allow patrons to contribute short video reviews.
3. Growth Pulse Dashboard: Provide vendors with deep-dive analytics on daily revenue and order volume trends. The Analytics Agent will correlate revenue data with weather and local events to suggest inventory adjustments.

Built With

• chart.js
• frontend:-react-19
• gemini-2.5-flash-native-audio)-mapping:-leaflet.js-with-customized-voyager-dark/inverted-styling.-visuals:-three.js-(3d-neural-coordination-mesh)
• google:aistudio
• spatial
• tailwind-css-intelligence:-@google/genai-(gemini-3-pro/flash-preview
• typescript