Trading Daily — Automated Pre-Market Intelligence Platform

Trading Daily is an automated financial intelligence system that delivers actionable pre-market briefings to traders every weekday morning. The platform orchestrates multiple data-collection agents, synthesizes market insights using Azure AI Foundry (GPT-4o), and distributes professional briefings via email all before the U.S. market opens.

---

Live Azure Deployment

The application is live and accessible at:

🔗 https://polite-pond-0a06a4e0f.1.azurestaticapps.net/

---

Problem Statement

Traders spend significant time each morning gathering data from multiple sources i.e economic calendars, sentiment indicators, earnings reports, and technical charts before they can form a trading plan. This manual process is time consuming, inconsistent, and prone to missing critical information.

Trading Daily solves this by automating the entire pre-market research workflow. Subscribers receive a single, AI analyzed briefing covering all major market factors at 5:30 AM EST, giving them a comprehensive view of the trading day ahead without switching between multiple platforms.

---

Features and Functionality

Multi-Agent Data Collection

Four specialized agents run concurrently every weekday morning to gather real time market data:

• Forex Factory Agent — Scrapes high impact USD economic events from the ForexFactory calendar, extracting event times, forecasts, actual values, and previous readings.
• Fear and Greed Agent — Captures the CNN Fear and Greed Index score and historical comparisons (previous close, 1 week, 1 month, 1 year ago) to gauge market sentiment.
• Earnings Agent — Collects upcoming earnings reports from Market Chameleon, including ticker symbols, report timing (pre-market/after-close), expected move percentages, and implied volatility.
• SPY Chart Agent — Fetches weekly SPY candle data from TradingView, calculates the 20-week Simple Moving Average, and determines the current trend position relative to the SMA.

AI-Powered Analysis

All collected data is fed into Azure AI Foundry (GPT-4o) which generates a structured briefing covering:

• Market structure assessment (SPY vs. 20-week SMA trend)
• Market sentiment summary based on the Fear and Greed Index
• Key economic events to watch during the session
• Notable earnings highlights and their potential impact
• Actionable trading considerations
• Overall risk assessment (Low / Moderate / High / Extreme)

Server-Side Chart Rendering

The platform renders SPY weekly price charts with SMA overlays as PNG images using Chart.js on the server. These charts are embedded directly in the email briefings as visual aids.

Subscription Management

A full subscription lifecycle is supported through the web interface:

• Email-based subscription with confirmation flow
• Topic selection for different market sessions (9:45 AM and 10:00 AM)
• Subscription status checking
• Unsubscribe with re-subscribe support

Scheduled Delivery

A cron-based scheduler triggers the full pipeline at 5:30 AM EST, Monday through Friday.

Responsive Web Interface

The frontend provides a clean subscription management portal with:

• Tabbed interface for subscribe, check status, and unsubscribe flows
• Light and dark mode with system preference detection and manual toggle
• Mobile-responsive layout

Light Mode	Dark Mode

---

Real-World Use Cases

Independent Day Traders — Receive a consolidated pre-market briefing instead of manually checking five or more websites each morning. The AI generated risk assessment helps prioritize which sessions to trade actively.

Swing Traders — The weekly SPY chart analysis with SMA trend positioning provides a macro context for multi-day trade decisions, delivered consistently without manual chart review.

Trading Communities and Educators — Newsletter operators can use Trading Daily as a content backbone, distributing professional-grade market analysis to their audience on a consistent schedule.

Portfolio Managers — Stay informed of high-impact economic events and earnings that could affect holdings, with sentiment context from the Fear and Greed Index.

---

Technology Stack

Microsoft and Azure Services

Service	Purpose
Azure AI Foundry (GPT-4o)	Analyzes aggregated market data and generates structured pre-market briefings with risk assessments
Azure Table Storage	Stores and manages subscriber records, confirmation tokens, subscription status, and metadata
Azure Blob Storage	Hosts chart snapshot images used as email attachments
Azure Container Apps	Hosts the backend API and scheduled agent pipeline with Managed Identity support
Azure Static Web Apps	Hosts the frontend subscription portal with SPA routing and security headers
Azure Identity (DefaultAzureCredential)	Provides credential management across Azure services using Managed Identity in production
Azure Communication Services (Email)	Sends transactional emails (subscription confirmations and pre-market briefings) via ACS Email

Frontend

Technology	Purpose
React 19	UI framework
TypeScript	Type-safe development
Vite 7	Build tooling and development server
Tailwind CSS 4	Utility-first styling
TanStack React Query	Server state management and data fetching
Axios	HTTP client for API communication
React Toastify	User notifications

Backend

Technology	Purpose
Express.js 5	REST API framework
TypeScript	Type-safe server development
Node-Cron	Scheduled task execution
Playwright	Headless browser automation for web scraping agents
Azure Communication Services SDK	Email delivery via ACS Email
Chart.js + chartjs-node-canvas	Server-side chart rendering to PNG
express-rate-limit	API rate limiting
Docker	Containerized deployment

Azure SDKs

• @azure-rest/ai-inference — Azure AI Foundry model inference
• @azure/communication-email — Azure Communication Services email sending
• @azure/data-tables — Azure Table Storage operations
• @azure/identity — Managed Identity and credential handling
• @azure/core-auth — Shared authentication utilities

---

Project Structure

trading-news/
├── backend/
│   ├── src/
│   │   ├── agents/              # Data collection agents
│   │   │   ├── forexFactoryAgent.ts
│   │   │   ├── fearGreedAgent.ts
│   │   │   ├── earningsAgent.ts
│   │   │   ├── spyChartAgent.ts
│   │   │   └── types.ts
│   │   ├── services/            # Core business logic
│   │   │   ├── agentOrchestratorService.ts
│   │   │   ├── foundryService.ts
│   │   │   ├── schedulerService.ts
│   │   │   ├── emailService.ts
│   │   │   ├── tableStorageService.ts
│   │   │   ├── blobStorageService.ts
│   │   │   └── chartRenderer.ts
│   │   ├── routes/              # API route definitions
│   │   ├── controllers/         # Request handlers
│   │   ├── middleware/          # Auth and rate limiting
│   │   ├── templates/           # Email HTML templates
│   │   └── types/               # Shared TypeScript types
│   ├── Dockerfile
│   └── package.json
├── frontend/
│   ├── src/
│   │   ├── components/          # UI components
│   │   ├── contexts/            # React context providers
│   │   ├── hooks/               # Custom React hooks
│   │   ├── api/                 # API client and endpoints
│   │   ├── types/               # TypeScript type definitions
│   │   └── utils/               # Utility functions
│   ├── staticwebapp.config.json
│   └── package.json
└── README.md

---

API Endpoints

Agent Routes (/api/v1/agents)

Method	Endpoint	Description
POST	/agents/run	Trigger pre-market briefing on-demand
GET	/agents/status	Check scheduler status

Subscription Routes (/api/v1)

Method	Endpoint	Description
POST	/subscribe	Create new subscription
POST	/subscription/confirm	Confirm subscription with token
GET	/subscription/:email	Get subscription details
DELETE	/subscription/:email	Remove subscription
GET	/subscriptions	List subscriptions by status

Email Routes (/api/v1)

Method	Endpoint	Description
POST	/send-mail	Send chart snapshots to subscribers

All routes (except public confirmation) require API key authentication via the x-api-key header.

---

Getting Started

Prerequisites

• Node.js 20+
• npm
• Azure account with the following resources provisioned:
  • Azure AI Foundry project with GPT-4o deployment
  • Azure Storage account (Table Storage and Blob Storage)
  • Azure Container Apps (for backend deployment)
  • Azure Static Web Apps (for frontend deployment)
• Azure Communication Services resource (for email)

Backend Setup

cd backend
npm install
npx playwright install chromium --with-deps
cp .env.example .env
# Fill in environment variables
npm run dev

Frontend Setup

cd frontend
npm install
cp .env.example .env.local
# Set VITE_API_URL and VITE_API_KEY (local dev values)
npm run dev

---

Environment Variables

Backend

Variable	Description
ACS_CONNECTION_STRING	Azure Communication Services connection string
ACS_SENDER_ADDRESS	ACS MailFrom address (e.g. DoNotReply@<guid>.azurecomm.net)
BLOB_STORAGE_BASE_URL	Azure Blob Storage base URL
AZURE_STORAGE_ACCOUNT_NAME	Azure Storage account name
AZURE_STORAGE_ACCOUNT_KEY	Storage account key (local dev only)
FOUNDRY_PROJECT_ENDPOINT	Azure AI Foundry project endpoint
FOUNDRY_MODEL_DEPLOYMENT_NAME	Model deployment name (e.g. gpt-4o)
FOUNDRY_API_KEY	Azure AI Foundry API key
SEND_MAIL_API_KEY	API key for route authentication
CORS_ORIGINS	Comma-separated allowed origins (omit to allow all)
FRONTEND_URL	Frontend URL for confirmation redirects
API_URL	Public backend URL (used in confirmation emails)
ENABLE_SCHEDULER	Enable/disable cron scheduler

Frontend

Variable	Description	File
VITE_API_URL	Backend API base URL	.env.local / .env.production
VITE_API_KEY	API key for authenticated requests	.env.local / .env.production

Use .env.local for local dev values and .env.production for production values. vite dev reads .env.local; vite build reads .env.production. Neither file is committed. See frontend/.env.example for the full template.

---

Team Members

Name	Microsoft Learn Username
Gerardo Arevalo	GerardoArevalo-3702
David Segun	DavidSegun-0630

---

Demo Video

https://youtu.be/gYEbMF0FkvM

---

Architecture Diagram

---

Email Screenshots

Pre-Market Briefing (5:30 AM)

Opening Range Bias (9:45 AM / 10:00 AM)