Trend Hunter

Trend Hunter is an AI-powered real-time trend-to-content engine that converts emerging social trends into ready-to-shoot entrepreneurial video scripts for Kenyan founders.

Problem

Kenyan creator teams often miss viral windows because they find trends too late, lack local context, and still have to decide the angle, hook, and script manually. The gap is not just discovery. It is turning a signal into a publishable short-form video idea inside the 24-72 hour window.

Target Users

• Kenyan short-form creators on TikTok, Reels, and YouTube Shorts covering entrepreneurship, money, side hustles, and founder culture.
• Early-stage startup marketing teams in East Africa.
• Media pages covering the African startup ecosystem.

V1 Features

1. Real-Time Trend Feed Aggregates social and Kenyan news signals into an emerging trend queue with velocity, category, and lifespan estimates.

2. AI Content Brief Generator Turns a selected trend into why it is happening, why it matters for Kenyan founders, a hook, a 30-60 second script, and a remix template.

3. Avalanche Trend Registry Registers verified trend events on Avalanche Fuji with a trend hash, timestamp, category, score snapshot, brief hash, and contributor.

Backend Architecture

The backend converts raw social signals into ranked trends, structured content briefs, and optional on-chain records.

Scrapers / APIs
  X | Reddit | Kenyan News | TikTok | YouTube
        |
Ingestion Layer
  normalization, deduplication, event queue
        |
Trend Engine
  clustering, velocity scoring, lifecycle stage
        |
AI Layer
  classification, local context, hook, script, remix template
        |
FastAPI
  trends, briefs, registry endpoints
        |
Avalanche C-Chain Service
  trend hash registration and verification

Implemented MVP modules:

• backend/app/ingestion/ collects normalized demo signals for X, Reddit, and Kenyan news.
• backend/app/trend_engine/ clusters signals and scores trends with velocity, cross-platform presence, and acceleration.
• backend/app/classifier/ classifies Founder Culture, Business, Money, or Not Relevant with deterministic MVP logic.
• backend/app/ai_brief_generator/ generates founder-focused ready-to-shoot briefs.
• backend/app/api is represented by backend/app/main.py, the FastAPI entrypoint.
• backend/app/blockchain/ hashes trends and submits to Avalanche Fuji when registry credentials are configured.
• backend/app/cache_layer/ provides a local TTL cache for hot trend leaderboards.

The production design still supports PostgreSQL, Redis, OpenAI or Claude, Celery/RQ, Playwright, Reddit API, X API, and RSS ingestion. The current 7-day MVP keeps those seams explicit while using deterministic local data so the demo works immediately.

API Endpoints

• GET /health
• POST /ingest/demo
• GET /trends
• GET /trends/{trend_id}
• POST /generate-brief with { "trend_id": "ai-interns" }
• POST /register-trend with { "trend_id": "ai-interns" }
• GET /registry

Avalanche Integration

Avalanche C-Chain is the verifiable Trend Registry layer. The smart contract records:

• trendHash
• firstSeen
• category
• score
• briefHash
• contributor

This creates transparent proof of who detected a trend first, prevents duplicate trend claims, and sets up a future contributor reputation or USDC rewards layer.

The static MVP uses an injected Core Wallet compatible provider for the Stage 1 demo. The product direction is an embedded wallet flow so non-crypto-native creators do not need to manage seed phrases or wallet setup.

Stage 1 Demo Flow

1. Open the dashboard and view the Emerging Trends feed.
2. Select the trend: AI agent replacing interns.
3. Click Generate Content Brief.
4. Review the founder-specific hook, script, local relevance metrics, and remix template.
5. Connect Core Wallet on Avalanche Fuji.
6. Paste the deployed TrendRegistry contract address.
7. Click Register on Avalanche.
8. The transaction log shows the Fuji Explorer transaction link.

Success means a user can move from trend discovery to publishable script to on-chain verified trend record.

Project Files

• contracts/YourContract.sol contains the TrendRegistry Solidity contract.
• test/YourContract.test.js covers registration, duplicate prevention, score updates, brief attachment, verification, trend hashes, and contributor reputation.
• frontend/index.html contains the mobile-first dashboard markup.
• frontend/styles.css contains dashboard styling.
• frontend/js/ contains split frontend logic: state.js, mock-data.js, api.js, trends.js, wallet.js, registry.js, assistant.js, toast.js, and init.js.
• scripts/deploy.js deploys TrendRegistry.

Run Locally

Install dependencies:

npm install

Install backend dependencies:

python -m pip install -r requirements-backend.txt

Compile:

npm run compile

Test:

npm test

Open the frontend:

xdg-open frontend/index.html

Run the backend:

npm run backend

Then open:

http://localhost:8000/docs

Deploy To Fuji

Create .env from .env.example and add a funded Fuji private key:

cp .env.example .env

Deploy:

npm run deploy:fuji

Paste the deployed contract address into the frontend, connect Core Wallet on Fuji, generate the brief, then register the trend on Avalanche.

For backend-side registration, set TREND_REGISTRY_ADDRESS and REGISTRY_PRIVATE_KEY in .env. Without those values, /register-trend returns a dry-run registry payload with the computed trend hash instead of submitting a transaction.