Agencio Predict

Login flows, MFA challenge, session management, user roles (8 roles with hierarchy), RBAC permissions, admin route protection, license validation, and subscription gates.

40+ external integrations across news, social, market data, prediction markets, derivatives, macro data, and ad platforms. Enrichment pipeline and scheduler orchestration.

DSL primitives (500+), NL→DSL translation, LLM Jury system, paper/live execution, guardrails (L1-L4), self-modification with cross-validation, and audit trails.

Bar-by-bar engine, 13 metrics (Sharpe, Sortino, Calmar, VaR/CVaR), walk-forward validation, Monte Carlo simulation, regime stress testing, and dataset management.

FVG detection, candlestick patterns, DTW template matching, Pattern Warming Radar, Pattern Hunter, regime-weighted confidence, and Wilson score intervals.

Physics-Informed Neural Networks with embedded trading constraints (Kelly, liquidation, mean-reversion). 16 inputs, 3 output heads, regime-aware training, S3 persistence.

AI billing (subscription/credits/BYOK), LLM routing, entity extraction, sentiment analysis, Stock Hunter AI, research reports, and cost control mechanisms.

417 trading primitives across 14 categories. Visual strategy builder, primitive browser, integration with AI Trader Wizard and Smart Portfolio. User flow diagrams.

Super admin unified trading view: active algorithm runs, mock/paper/live trades, guardrail kills, AI usage, broker latency tracking, and trade source classification.

Autonomous multi-strategy fund management: 5 brokers, 6 asset classes, 6 regime strategies, automatic strategy switching, NAV tracking, multi-symbol execution, and rebalancing.

Market regime detection (BULL/BEAR/SIDEWAYS/CRISIS/RECOVERY) using PINN + Markov HMM, automatic strategy switching, fitness scoring, and configurable thresholds.

Unified AI orchestration: 16 data source categories, consensus engine, black swan detection, executor kill switch integration, and real-time data aggregation.

Crisis detection: 12 event types, 4 severity levels, yield curve inversion, credit spreads, intermarket divergence, historical crisis matching, and executor integration.

Prediction market fraud detection: CLOB API, Union-Find clustering, 5 connection types, 10 DSL primitives, D3.js visualization, and manipulation risk scoring.

7 detection types (spoofing, wash trading, pump & dump, stop-hunt), 22 DSL primitives, L2 order book analysis, tick classification, and executor circuit breaker integration.

Human vs automation composite scoring, divergence engine (5 states), arb-bot detection, trust layer integration, and anomaly signal generation.

Competitive analysis, pricing tiers ($29-$199/mo), white-label licensing ($50K-$5M), claim validation against code, quantitative infrastructure documentation.

Honest platform assessment: pre-production readiness, API key requirements, paper trading plan, roadmap, known issues, and pre-production checklist.

Comprehensive guide explaining trading concepts: derivatives, options, futures, forex, crypto, technical analysis, risk metrics, microstructure, FIX protocol, and how they relate to the platform.

Complete DSL primitive reference: 506 primitives across 16 categories. Technical indicators, PINN ML, factor models, options Greeks, microstructure, execution algorithms, and more.

Executive summary: "Institutional Quant. Retail Accessible." 506 primitives, 16 categories, 5 brokers, 40+ integrations. Platform pillars, quant highlights, and technical foundation overview.

Quick start guide for new users. Account setup, platform navigation, and first steps with the prediction terminal.

How to use AI-powered trading features including algorithm creation, backtesting, paper trading, and live execution.

Connect your brokerage accounts (Alpaca, Binance, IBKR, Pepperstone, Schwab) for paper and live trading execution.

Claude-powered stock research with 39 presets across 18 categories. Generate thesis, risk assessment, and price targets.

Unified AI orchestration layer combining prediction markets, market data, sentiment, and macro signals into probability views.

Complete index of all 23 user guide chapters covering every feature of the platform.

Monorepo structure, package layout, deployment architecture (Vercel + AWS EC2), database schemas, and service boundaries.

7 schemas (auth, predict, marketing, trading, derivatives, omniscient, billing) with 28+ tables and migration conventions.

430+ API endpoints with authentication requirements, rate limits, request/response schemas, and usage examples.

DSL specification, NL→DSL translation, LLM Jury system, anti-hallucination defenses, and autonomous trading hardening.

5 broker adapters (Alpaca, Binance, IBKR, Pepperstone, Schwab), OAuth flows, credential encryption, and execution modes.

Super admin real-time monitoring of all trading activity: algorithm runs, mock/paper/live trades, guardrail kills, AI usage, and broker latency metrics.

Complete searchable index of all 100+ documentation files with classification badges and descriptions.

Comprehensive compliance and audit dossier for the AI trading system covering regulatory requirements and controls.

Full audit compliance documentation for platform review and regulatory submissions.

Mathematical audit of trading algorithms, risk metrics, and quantitative methodologies.

Integration guide for prediction market data sources including Polymarket, Kalshi, and Metaculus.

Non-disclosure agreement template for individual contractors and testers.

Open the documentation viewer to browse all formatted PDFs and Word documents.

Comprehensive proof, validation, and backtest testing. 7 test suites covering 50+ individual tests. Validates mathematical accuracy, backtesting correctness, and end-to-end feature workflows.

12 institutional-grade quant modules: cointegration, Fama-French, GARCH, lead-lag, momentum, TCA, execution algorithms, structural breaks, and sector rotation. 63 DSL primitives.

Expert-level mathematical audit of trading systems. 27 enhancement recommendations, 11 implemented. Covers Sharpe/Sortino, VaR/CVaR, GARCH, Almgren-Chriss, Kelly criterion.

Interactive Mermaid diagrams showing how SILO Cortex bolts into Predict: architecture, enrollment, event flow, TDS scoring, circuit breakers, and kill switch integration.

Overview of the SILO institutional integration, architecture, and partnership details.

Technical implementation details, API contracts, authentication flows, and data exchange protocols.

Detailed specification document covering endpoints, schemas, error handling, and deployment requirements.