Maps news events to price candlesticks to display market-response correlation. Crypto news analysis pipeline that ingests news, deduplicates via semantic vectors, analyzes sentiment with AI agents, snapshots price, and tracks realized price movements over time. NOTE: This MVP is 100% vibe-coded and support BTC only.
- Ingests news from MarketAux API for configured crypto tickers (e.g. BTC)
- Deduplicates using cosine similarity on embeddings — both within the current batch and against the last 24h in the vector DB
- Analyzes & filters via a LangGraph
StateGraph(conditional routing):- Junior analyst (GPT-5 Nano) evaluates each news item in parallel
- Noise/unimportant news is discarded (never stored) — only price-driving news is kept
- If confidence < threshold → graph routes to Senior analyst (GPT-5 Mini)
- Snapshots price from Binance perpetual futures at ingestion time
- Stores enriched vectors in Qdrant Cloud with full metadata
- Backfills realized price deltas (1h, 24h, 7d, 30d) via a separate cron-triggered endpoint
| Component | Technology |
|---|---|
| API Framework | FastAPI |
| Agent Orchestration | LangGraph |
| Vector Database | Qdrant Cloud |
| LLM Provider | Azure AI (GPT-5 Nano / Mini) |
| Embeddings | Azure AI text-embedding-3-large (256 dims) |
| Price Data | Binance USDⓈ-M Futures API |
| News Source | MarketAux API |
| Hosting | Render (web service + cron jobs) |
| Language | Python 3.14 |
| Toolchain | uv |
┌─────────────────────────────────────────────────────────────────┐
│ Render Platform │
├─────────────────────────────────────────────────────────────────┤
│ │
│ ┌──────────────┐ ┌──────────────┐ │
│ │ Cron: 10 min │────▶│ POST │ │
│ │ read-news │ │ /api/read-news│ │
│ └──────────────┘ └──────┬───────┘ │
│ │ │
│ ┌──────────────┐ ┌──────┴────────┐ │
│ │ Cron: 15 min │────▶│ POST │ │
│ │ update-prices│ │/api/update- │ │
│ └──────────────┘ │ prices │ │
│ └──────┬────────┘ │
│ │ │
│ ┌────────────────────┼────────────────────┐ │
│ │ FastAPI App │ │
│ │ │ │
│ │ news_pipeline/ │ price_updater/ │ │
│ │ ┌─────────────┐ │ ┌─────────────┐ │ │
│ │ │ fetch_news │ │ │ updater │ │ │
│ │ │ dedup │ │ │ (delta calc)│ │ │
│ │ │ graph (LG) │ │ └──────┬──────┘ │ │
│ │ │ agents │ │ │ │ │
│ │ │ store │ │ │ │ │
│ │ └──────┬──────┘ │ │ │ │
│ │ │ │ │ │ │
│ │ shared/│ │ │ │ │
│ │ ┌──────┴──────────┴──────────┴──────┐ │ │
│ │ │ qdrant.py embeddings.py binance │ │ │
│ │ └──────────────────────────────────┘ │ │
│ └─────────────────────────────────────────┘ │
└─────────────────────────────────────────────────────────────────┘
│ │ │
▼ ▼ ▼
┌─────────────┐ ┌──────────────┐ ┌─────────────┐
│ Qdrant Cloud│ │ Azure AI │ │ Binance │
│ (vectors + │ │ (embed + │ │ Futures │
│ metadata) │ │ LLM) │ │ API │
└─────────────┘ └──────────────┘ └─────────────┘
▲
│
┌─────────────┐
│ MarketAux │
│ (news feed) │
└─────────────┘
MarketAux API ──▶ Raw news articles (filtered by ticker)
│
▼
Extract full text from article URLs (trafilatura)
│
▼
Skip articles with unavailable URLs (403, 404, timeout → dropped from pipeline)
│
▼
Condense oversized articles (>2000 chars) via GPT-5 Nano summarization
│
▼
Embed all articles (Azure AI batch) ──▶ 256-dim vectors
│
▼
Intra-batch dedup (pairwise cosine ≥ 0.90 → keep first)
│
▼
RAG dedup vs Qdrant (last 24h, cosine ≥ 0.90 → discard)
│
▼
Attach similar past news as context (with realized price data)
│
▼
LangGraph StateGraph (graph.py):
┌─────────────────────────────────────────────────────────┐
│ START → junior_analyst (GPT-5 Nano) │
│ │ │
│ ├── discard=true → END (news dropped) │
│ ├── confidence ≥ 0.75 → END (result kept) │
│ └── confidence < 0.75 → senior_analyst │
│ (GPT-5 Mini) → END │
└─────────────────────────────────────────────────────────┘
Batch: asyncio.gather over graph.ainvoke() per news item
│
▼
Fetch Binance mark price (BTCUSDT perpetual) — only for kept news
│
▼
Batch upsert to Qdrant (vector + full metadata payload)
Query Qdrant: news with null realized_price_delta fields
where published_at is old enough for each delta window
│
▼
Fetch historical prices from Binance klines at:
published_at + 1h, +24h, +7d, +30d
│
▼
Calculate: (historical_price - price_at_ingestion) / price_at_ingestion
│
▼
Batch update Qdrant payloads with realized deltas
{
"ticker": "BTC",
"source": "coindesk",
"published_at": "2026-05-21T18:30:00Z",
"news_summary": "The FED cuts rates by 50 bps, signaling aggressive easing...",
"news_full_text": "Full article text...",
"sentiment": "bullish",
"impact": 3,
"confidence": 0.78,
"predicted_by_model": "gpt-5.4-nano",
"price_at_ingestion": 68250.00,
"realized_price_delta_pct_1h": null,
"realized_price_delta_pct_24h": null,
"realized_price_delta_pct_7d": null,
"realized_price_delta_pct_30d": null
}Sentiment: Binary — "bullish" (positive price pressure) or "bearish" (negative price pressure). No neutral — if news can't clearly drive price in either direction, it's discarded as noise and never stored.
Impact Scale (1-3):
| Level | Name | Timeframe | Examples |
|---|---|---|---|
| 1 | Notable | Moves price for hours | Coin listed on major exchange (Binance, Coinbase) · Significant token burn · Large whale movement · Successful protocol upgrade · Major partnership |
| 2 | High | Moves price for days | ETF approval/rejection · Major hack/exploit ($100M+) · Country bans/legalizes crypto · Institutional entry/exit (BlackRock, Fidelity) · Exchange withdrawal freeze |
| 3 | Extreme | Moves entire market for weeks | Fed rate decision · Major stablecoin depeg · Exchange collapse (FTX-level) · International regulatory crackdown · Global banking crisis |
All tunable parameters live here. No restart needed for threshold changes (reload on next request).
tickers: [BTC]
dedup:
cosine_threshold: 0.90 # similarity score to consider as duplicate
lookback_hours: 24 # how far back to check for duplicates
agents:
confidence_threshold: 0.75 # below this → escalate to senior
nano_deployment: gpt-5.4-nano
mini_deployment: gpt-5.4-mini
api_version: "2025-04-01-preview"
reasoning_effort: high
embeddings:
deployment: text-embedding-3-large
dimensions: 256
api_version: "2023-05-15"
pipeline:
fetch_news_interval_minutes: 10
max_full_text_chars: 2000 # articles longer than this get LLM-summarized
price_updater:
deltas: [1h, 24h, 7d, 30d]MARKETAUX_API_KEY=...
QDRANT_URL=https://your-cluster.cloud.qdrant.io
QDRANT_API_KEY=...
AZURE_AI_ENDPOINT=https://your-resource.cognitiveservices.azure.com
AZURE_AI_API_KEY=...cd backend
# Install dependencies with uv
uv sync
# Copy and fill in secrets
cp .env.example .env
# Run the API server
uv run start
# Trigger pipelines manually
curl -X POST http://localhost:8000/api/read-news
curl -X POST http://localhost:8000/api/update-prices
# Run tests
uv run pytestcd frontend
# Install dependencies
npm install
# Copy env (defaults to localhost:8000 backend)
cp .env.example .env.local
# Run dev server (http://localhost:3000)
npm run dev
# Build static export
npm run build| Component | Technology |
|---|---|
| Framework | Next.js 16 (App Router, static export) |
| Language | TypeScript |
| Charts | TradingView Lightweight Charts v5 |
| Styling | Tailwind CSS v4 |
| Price Data | Binance USDⓈ-M Futures REST API (client-side) |
| News Data | Backend FastAPI /api/news endpoint |
┌─────────────────────────────────────────────────────────┐
│ Browser (SPA) │
│ │
│ ┌────────────────────────────────────────────────────┐ │
│ │ Toolbar: [Ticker] [1h|4h|1d|1w|1M] [Impact ━━━━] │ │
│ ├────────────────────────────────────────────────────┤ │
│ │ │ │
│ │ TradingView Lightweight Chart │ │
│ │ (candlestick + news markers) │ │
│ │ │ │
│ │ ● ● ● ● │ │
│ │ ║║ ║║ ║║ ║║ ║║ ║║ ║║ │ │
│ │ ●║║ ║║ ║║ ●║║ ║║ ║║ ║║ │ │
│ │ │ │
│ └────────────────────────────────────────────────────┘ │
│ │
│ ┌──────────────────────┐ │
│ │ Popup (on click): │ │
│ │ • 🟢 BTC surges... │ │
│ │ • 🔴 SEC warns... │ │
│ └──────────────────────┘ │
└──────────────────────────────────────────────────────────┘
│ │
▼ ▼
┌──────────────┐ ┌──────────────┐
│ Binance │ │ FastAPI │
│ fapi/v1/ │ │ /api/news │
│ klines │ │ (→ Qdrant) │
└──────────────┘ └──────────────┘
- User selects ticker (e.g.,
BTCUSDT) and timeframe (e.g.,1d) - Frontend fetches klines directly from Binance Futures public API
- Chart renders candlesticks and subscribes to visible range changes
- When visible range changes (scroll/zoom), frontend fetches news from backend for that time window
- News items are aggregated per candle:
- Color = sentiment (red = bearish, green = bullish)
- Size = impact (1 = smallest circle, 3 = largest)
- Impact slider filters out news below the threshold (client-side re-aggregation)
- Click on a marker shows popup with news items sorted by impact (most impactful first)
GET /api/news?ticker=BTC&from_ts=2026-01-01T00:00:00Z&to_ts=2026-05-28T00:00:00Z&min_impact=2
Response:
{
"items": [
{
"published_at": "2026-05-21T18:30:00Z",
"sentiment": "bullish",
"impact": 3,
"news_summary": "BTC ETF inflows hit record...",
"confidence": 0.82,
"predicted_by_model": "gpt-5.4-nano",
"price_at_ingestion": 68250.00
}
],
"count": 42
}# Navigate to frontend
cd frontend
# Install dependencies
npm install
# Copy env file and configure API URL
cp .env.example .env.local
# Run dev server
npm run dev
# → http://localhost:3000
# Build for production (static export)
npm run build
# Output: frontend/out/frontend/
├── src/
│ ├── app/
│ │ ├── page.tsx # Main page (ticker, timeframe, impact state)
│ │ ├── layout.tsx # Root layout (light theme, monospace font)
│ │ └── globals.css # Tailwind + CSS variables
│ ├── components/
│ │ ├── Chart.tsx # TradingView chart + markers + range subscription
│ │ ├── Toolbar.tsx # Ticker input, timeframe buttons, impact slider
│ │ └── NewsPopup.tsx # Hover/click popup with news bullet list
│ └── lib/
│ ├── types.ts # Shared TypeScript types
│ ├── klines.ts # Binance futures klines fetcher
│ ├── news.ts # Backend news API fetcher
│ └── aggregate.ts # News aggregation, sentiment colors, impact sizing
├── next.config.ts # Static export config
├── .env.example # NEXT_PUBLIC_API_URL template
└── package.json
The render.yaml at the repo root defines:
- Web service: FastAPI app serving pipeline + chart API endpoints
- Static site: Next.js frontend (static export to
out/) - Cron job (15 min): hits
/api/read-news - Cron job (60 min): hits
/api/update-prices
Set all environment variables in the Render dashboard:
- Backend:
MARKETAUX_API_KEY,QDRANT_URL,QDRANT_API_KEY,AZURE_AI_ENDPOINT,AZURE_AI_API_KEY - Frontend:
NEXT_PUBLIC_API_URL(set to the backend service's external URL)
The backend reads config.yaml from the repo for non-secret settings.