🏙️ FUNDR — AI-Powered Local Business Insights

FUNDR helps banks, investors, and cities discover and fund thriving small businesses using real-world data from Google Local and AI-driven analytics.

Built with FastAPI, SerpAPI, and ChatGPT, FUNDR transforms raw Google Local data into actionable leaderboards, insights, and impact scores — highlighting the local businesses most deserving of funding.

---

🚀 Key Features

• 🔍 Discover Local Businesses
  • Search by city and category using Google Local data.
• 🧮 FUNDR Score (0–100)
  • Quality – Ratings + Review Volume
  • Consistency – Recency + Rating Trend
  • Community – Positive Sentiment (clean, kind, inclusive, affordable)
  • Ops Fit – Operational completeness (hours, phone, website)
• 🏆 Leaderboard Mode
  • Ranks top businesses per city and category.
• ⚡ Quota-Efficient Design
  • Only 7 SerpAPI calls per full leaderboard (1 discovery + 3×[details + reviews]).
• 💬 AI Insights
  • Generates “Why Fund,” “Top Praise,” and “Top Fix” summaries per business.

---

🧩 Architecture Overview

User → FUNDR FastAPI → SerpAPI (Google Local + Reviews)
                          ↓
             Scoring Engine (Quality, Consistency, Community, OpsFit)
                          ↓
                 Cache + Leaderboard → JSON / OpenAPI / Custom GPT

🛠️ Tech Stack

Layer Technology
Framework FastAPI (Python)
Data Source SerpAPI (Google Local + Reviews)
AI Integration ChatGPT / Custom GPT Action
Caching In-memory TTL (6 hours)
Scoring FUNDR algorithm (Quality • Consistency • Community • Ops Fit)

🔗 API Endpoints

Endpoint	Method	Description
/	GET	API root information
/discover	GET	Discover small businesses by city and category
/business/{place_id}	GET	Retrieve detailed info + reviews for one business
/leaderboard	GET	Get ranked businesses with FUNDR scores
/score/recompute	POST	Recompute scores for all cached businesses

🧮 Example Usage

Request

curl "http://127.0.0.1:8000/leaderboard?city=Detroit&category=coffee%20shops&limit=10"

Response

{
  "city": "Detroit",
  "category": "coffee shops",
  "updated_at": "2025-10-17T19:00:00Z",
  "items": [
    {
      "place_id": "ChIJd8BlQ2BZwokRAFUEcm_qrcA",
      "name": "The Congregation Detroit",
      "fundr_score": 88.6,
      "subscores": {
        "quality": 0.92,
        "consistency": 0.88,
        "community": 0.75,
        "ops_fit": 0.9
      },
      "why_fund": "Strong praise for friendly service and community feel; opportunity to improve parking.",
      "badges": ["Community Favorite"]
    }
  ]
}

👤 Author

Jaylen Thomas
Software Engineer II @ JPMorgan Chase
Graduate Student (AI & Data Science)
Detroit-native • Climate-Tech • Community Impact Builder

🧠 Example “Why Fund” Insight

“Strong recent customer praise for friendly staff and community engagement; opportunity to improve wait times.”

⚡ Performance & Quota Protection

Each leaderboard request capped at 7 SerpAPI calls

Cached results live for 6 hours

Chain stores auto-filtered (200+ national brands)

Repeated queries = 0–1 API calls

🧩 Integrating with Custom GPT

FUNDR’s OpenAPI schema lets you build an interactive Custom GPT.

Example Prompt:

“Find the top woman-owned bakeries in Atlanta.”

GPT Flow:

/discover → Fetch candidates

/leaderboard → Rank + explain “Why Fund”

Output → Contextual leaderboard and AI insights

🧭 Vision

FUNDR empowers cities, banks, and communities to invest in small businesses that uplift local economies — using data, not guesswork.

🏃 Running the App

uvicorn app.main:app --reload