Power User

Source: Notion | Last edited: 2026-01-28 | ID: 2f62d2dc-3ef...

Alpha Forge Platform - Product Requirements Document (PRD)

Overview

Product: Alpha Forge Platform - Strategy Management & Execution System

Vision: A unified platform where users can deploy, manage, backtest, paper trade, and eventually live trade their quantitative strategies.

Strategic Context

Two User Segments, Two Entry Points

The platform will ultimately serve two distinct user segments with different ways to add strategies:

┌─────────────────────────────────────────────────────────────────────────────┐ │ Platform Entry Points │ ├─────────────────────────────────────────────────────────────────────────────┤ │ │ │ POWER USERS (Technical) RETAIL USERS (Non-Technical) │ │ ┌─────────────────────────┐ ┌─────────────────────────┐ │ │ │ │ │ │ │ │ │ alpha-forge (Local) │ │ Natural Language Chat │ │ │ │ DSL Development │ │ “Create a momentum │ │ │ │ │ │ │ strategy for BTC…” │ │ │ │ ▼ │ │ │ │ │ │ │ Deploy via CLI/API │ │ ▼ │ │ │ │ │ │ AI Generates DSL │ │ │ └───────────┬─────────────┘ └───────────┬─────────────┘ │ │ │ │ │ │ │ ┌─────────────┐ │ │ │ └────────▶│ PLATFORM │◀──────────┘ │ │ │ │ │ │ │ • Manage │ │ │ │ • Backtest │ │ │ │ • Paper │ │ │ │ • Live │ │ │ └─────────────┘ │ │ │ │ ════════════════════════════════════════════════════════════════════════ │ │ NOW: Build Power User path first (we are power users, deeper understanding)│ │ FUTURE: Add Retail User path (natural language → strategy) │ │ ════════════════════════════════════════════════════════════════════════ │ │ │ └─────────────────────────────────────────────────────────────────────────────┘

Why Power Users First? ┌──────────────────────┬──────────────────────────────────────────────────────────────────────┐ │ Reason │ Explanation │ ├──────────────────────┼──────────────────────────────────────────────────────────────────────┤ │ We are the users │ Team uses alpha-forge daily, deep understanding of needs │ ├──────────────────────┼──────────────────────────────────────────────────────────────────────┤ │ Clearer requirements │ Technical users have explicit needs, less ambiguity │ ├──────────────────────┼──────────────────────────────────────────────────────────────────────┤ │ Foundation first │ Core platform features (backtest, paper trade) needed for both paths │ ├──────────────────────┼──────────────────────────────────────────────────────────────────────┤ │ Faster validation │ Can dogfood immediately, iterate quickly │ └──────────────────────┴──────────────────────────────────────────────────────────────────────┘ Future Convergence

Once a strategy exists in the platform (regardless of how it was created), all downstream features are shared:

Viewing strategy content and history
Running backtests
Paper trading
Live trading
Performance monitoring This means building the Power User path first creates the foundation for Retail User path.

Target Users

Phase 1 Focus: Power Users (Technical Traders)

Developers and quants who write strategies in alpha-forge DSL
Comfortable with CLI, YAML, and technical concepts
Want to manage, backtest, and deploy strategies they’ve developed locally Future: Retail Users (Non-Technical Traders)
Traders who understand trading concepts but don’t code
Will use natural language to describe strategies
Platform + AI generates DSL on their behalf

User Journey (Power User)

┌─────────────────────────────────────────────────────────────────────────────┐ │ Power User Journey │ ├─────────────────────────────────────────────────────────────────────────────┤ │ │ │ 1. LOCAL DEVELOPMENT │ │ ┌──────────────────────────────────────────┐ │ │ │ $alpha_forge run my_strategy.yaml │ │ │ │ # iterate locally, test ideas │ │ │ └──────────────────────────────────────────┘ │ │ │ │ │ ▼ │ │ 2. DEPLOY TO PLATFORM │ │ ┌──────────────────────────────────────────┐ │ │ │$ alpha_forge deploy my_strategy.yaml │ │ │ │ # or upload via web UI │ │ │ └──────────────────────────────────────────┘ │ │ │ │ │ ▼ │ │ 3. MANAGE & ITERATE (Platform) │ │ ┌──────────────────────────────────────────┐ │ │ │ • View strategy list │ │ │ │ • Run backtests (different params) │ │ │ │ • Compare results │ │ │ │ • Update strategy versions │ │ │ └──────────────────────────────────────────┘ │ │ │ │ │ ▼ │ │ 4. PAPER TRADING │ │ ┌──────────────────────────────────────────┐ │ │ │ • Activate paper trading │ │ │ │ • Monitor signals and simulated P&L │ │ │ │ • Validate strategy in live conditions │ │ │ └──────────────────────────────────────────┘ │ │ │ │ │ ▼ │ │ 5. LIVE TRADING (Future) │ │ ┌──────────────────────────────────────────┐ │ │ │ • Connect broker │ │ │ │ • Deploy to live │ │ │ │ • Monitor real performance │ │ │ └──────────────────────────────────────────┘ │ │ │ └─────────────────────────────────────────────────────────────────────────────┘

Feature Requirements

Phase 1: Strategy Management (MVP) - Power Users

1.1 Strategy Deployment

Users can deploy strategies from local alpha-forge to platform
Methods:
- CLI: alpha_forge deploy <strategy.yaml> (preferred for power users)
- Web UI: Upload YAML file directly
Platform validates DSL before accepting
Support versioning (update existing strategy, view history) 1.2 Strategy Viewing
List all deployed strategies
View strategy details:
- DSL content (YAML, syntax highlighted)
- Metadata (name, description, version, dates)
- Summary (optional text description)
Search and filter strategies 1.3 Backtest History
View all historical backtests for a strategy
Each backtest shows:
- Summary metrics (Sharpe, drawdown, returns, etc.)
- Equity curve chart
- Trade list (detailed view)
- Parameters used (date range, symbols)
Compare across runs Phase 2: Execution

2.1 Training (for ML strategies)

Trigger training from platform
Track progress, view logs
Store trained model artifacts 2.2 Backtesting
Run backtests with custom parameters:
- Date range
- Symbol selection
- Strategy parameters
Async execution with progress updates
Results stored and viewable Phase 3: Paper Trading

3.1 Activation

Start paper trading for any strategy
Configure: symbols, position sizing 3.2 Monitoring
Real-time signals and fills
Simulated positions and P&L
Performance charts
Compare to backtest expectations Phase 4: Live Trading (Future)
Broker integration
Live deployment
Real-time monitoring
Risk controls

Future: Retail User Path

Note: This is NOT in current scope, documented for architectural awareness.

When retail user support is added:

Entry Point:

Chat interface: “Create a strategy that…”
AI generates DSL using Claude + alpha-forge context Integration:
Generated DSL is stored in platform (same as deployed strategies)
All downstream features work identically
User can view/edit generated DSL if desired Architectural Implication:
Strategy storage must be source-agnostic (deployed vs generated)
Consider source field: cli_deploy, web_upload, ai_generated

Technical Notes

Relationship to Existing Codebases: ┌──────────────────────┬──────────────────────────┬──────────────────────────────────────────────────────┐ │ Repository │ Purpose │ Role in This PRD │ ├──────────────────────┼──────────────────────────┼──────────────────────────────────────────────────────┤ │ alpha-forge │ Core quantitative engine │ Library dependency, provides DSL compiler, execution │ ├──────────────────────┼──────────────────────────┼──────────────────────────────────────────────────────┤ │ alpha-forge-platform │ User-facing web app │ Where these features are implemented │ └──────────────────────┴──────────────────────────┴──────────────────────────────────────────────────────┘ Key Integration Points:

DSL validation: alpha_forge.compiler.compile_dict()
Strategy execution: alpha_forge.api.run_strategy()
Signal generation: alpha_forge signal command

Success Metrics

┌───────────────────────────────────┬────────┬─────────┐ │ Metric │ Target │ Phase │ ├───────────────────────────────────┼────────┼─────────┤ │ Team members using platform daily │ 100% │ Phase 1 │ ├───────────────────────────────────┼────────┼─────────┤ │ Strategies deployed │ > 20 │ Phase 1 │ ├───────────────────────────────────┼────────┼─────────┤ │ Backtests run per week │ > 50 │ Phase 2 │ ├───────────────────────────────────┼────────┼─────────┤ │ Paper trading sessions active │ > 5 │ Phase 3 │ └───────────────────────────────────┴────────┴─────────┘

Timeline

┌─────────┬─────────────────────────────────────────────┬──────────┐ │ Phase │ Scope │ Target │ ├─────────┼─────────────────────────────────────────────┼──────────┤ │ Phase 1 │ Strategy Management (deploy, view, history) │ Q1 2026 │ ├─────────┼─────────────────────────────────────────────┼──────────┤ │ Phase 2 │ Execution (train, backtest) │ Q2 2026 │ ├─────────┼─────────────────────────────────────────────┼──────────┤ │ Phase 3 │ Paper Trading │ Q3 2026 │ ├─────────┼─────────────────────────────────────────────┼──────────┤ │ Phase 4 │ Live Trading + Broker Integration │ Q4 2026+ │ ├─────────┼─────────────────────────────────────────────┼──────────┤ │ Future │ Retail User Natural Language Entry │ TBD │ └─────────┴─────────────────────────────────────────────┴──────────┘