Product Value Overview

Competitive positioning, pricing guidance, and claim validation

ā† Back to Documentation Hub

THE SIGNAL FABRIC

Decision intelligence that fuses prediction markets, derivatives, macro, sentiment, and price action
One platform. Every signal. Every team.

šŸ“Š Platform Statistics VERIFIED

143+
DSL Primitives
32+
Data Integrations
5
Broker Adapters
270+
DB Migrations
430+
API Endpoints
2,600
Dev Hours
$417K
Replacement Cost
39
Stock Hunter Presets
Verification Sources

Statistics verified against: CLAUDE.md (project index), COMPLETED.md (work log), packages/be/src/algorithms/dsl/types.ts (primitives), db/migrations/ (schema count)

šŸ’° Competitive Pricing Analysis

Competitor Price AI Algo Builder Multi-Broker Prediction Markets Email Intelligence
Agencio Predict $79-199/mo āœ“ 143 primitives āœ“ 5 brokers āœ“ 4 markets āœ“ Unique
TradingView $15-60/mo āœ— Pine Script only āœ— Partner links āœ— None āœ— None
QuantConnect $20-200/mo āœ— Manual coding ~ Via partners āœ— None āœ— None
Trade Ideas $118-228/mo āœ— Scanners only āœ— None āœ— None āœ— None
Composer $15-99/mo ~ Visual builder āœ— Alpaca only āœ— None āœ— None
Bloomberg $24,000/yr āœ— BQL scripting āœ— Via partners āœ— None āœ— None
Value Stack Analysis

A serious retail trader today pays: TradingView Pro+ ($60) + QuantConnect ($100) + Trade Ideas ($200) + data feeds ($100) = $460/mo

Agencio Predict replaces 80% of this stack at $79-199/mo.

šŸ·ļø Recommended Pricing Tiers

Free
$0
Testing & Evaluation
  • Mock trading ($100K simulated)
  • 3 algorithms max
  • BYOK only (no AI credits)
  • Basic signals
Starter
$29/mo
Hobbyists
  • Paper trading
  • 5 algorithms
  • 25 AI credits/month
  • Basic Stock Hunter
Trader
$79/mo
Active Retail
  • Live trading (all brokers)
  • 15 algorithms
  • 100 AI credits/month
  • Full Stock Hunter
  • Terminal access
Pro
$199/mo
Serious Traders
  • Unlimited algorithms
  • 300 AI credits/month
  • AI Fund Manager
  • Institutional data (13F/13D)
  • Priority support

šŸ¢ White-Label & Licensing Options

License Tier Annual Fee Target Buyer What's Included
SaaS Reseller $50K-100K + 15-25% rev share Fintechs, neobrokers, communities White-label UI, we host, up to 500 users, custom domain
Self-Hosted $150K-300K/year Hedge funds, family offices, regulated entities Docker containers, your infrastructure, SSO, audit logs
Source License $500K-1M one-time Strategic acquirers, platform companies Full source code, git history, documentation, 2-week KT
Platform Acquisition $2M-5M PE firms, competitors, strategic buyers All IP, brand, code, customer relationships
ROI Analysis for White-Label Buyers

Neobroker adding signal intelligence:

ā€¢ Build internally: $500K + 24 months + ongoing team

ā€¢ License Agencio: $100K/year + 4 weeks to launch

ā€¢ Break-even: Month 3 (assuming 500 users @ $79/mo)

šŸŽÆ Unique Differentiators

1. Email-to-Trading-Signals Pipeline

Convert newsletter intelligence into DSL trading primitives. Nobody else does this.

šŸ“ packages/be/src/mailbox/insights/

2. Human vs. Bot Classification

5-state divergence engine classifies market moves as human-driven vs bot-driven.

šŸ“ packages/be/src/insights/divergence-engine.ts

3. Prediction Market Integration

Polymarket, Kalshi, Metaculus as first-class DSL primitives with arbitrage detection.

šŸ“ packages/be/src/integrations/polymarket.ts

4. Plain English ā†’ Trading Algorithm

NL ā†’ DSL with 9 anti-hallucination defenses, LLM Jury, cross-validation.

šŸ“ packages/be/src/algorithms/llm/

āœ… Claim Validation ā€” Sanity Check

Every marketing claim verified against shipping code. Green = verified in codebase. Amber = partially implemented. Red = not yet built.

āœ“
Verified
"143+ DSL signal primitives for algorithm building"
Counted in PRIMITIVES registry. Includes: price (15), technical (22), sentiment (18), macro (12), prediction_market (8), derivatives (14), institutional (13), mailbox (8), plus sizing, position, options, fundamental categories.
šŸ“ packages/be/src/algorithms/dsl/types.ts:PRIMITIVES
āœ“
Verified
"5 broker integrations (Alpaca, Binance, IBKR, Pepperstone, Schwab)"
All 5 adapters implement BrokerAdapter interface. Alpaca/Binance: API key auth. IBKR: session-based. Pepperstone/Schwab: OAuth 2.0 with PKCE.
šŸ“ packages/be/src/brokers/{alpaca,binance,ibkr,pepperstone,schwab}.ts
āœ“
Verified
"PINN ML predictor with trading constraints embedded in loss function"
TensorFlow.js model with 16 inputs, 3 outputs (price, velocity, liquidation risk). Kelly sizing, liquidation avoidance, mean-reversion bounds in loss. S3 persistence, graduation gates.
šŸ“ packages/be/src/predictions/pinn-predictor.ts
āœ“
Verified
"Almgren-Chriss market impact model for trade execution"
Square-root market impact formula implemented. Bid-ask spread tiers by asset class. Mode-aware slippage calculation (backtest/paper/live).
šŸ“ packages/be/src/trading/services/trade-execution.ts
āœ“
Verified
"VPIN and Kyle's Lambda tick-level analysis"
Volume-Synchronized Probability of Informed Trading and price impact coefficient implemented. Consumes Binance aggTrades (crypto) and Polygon trades (equities).
šŸ“ packages/be/src/insights/tick-classifier.ts
āœ“
Verified
"13 backtest metrics including Cornish-Fisher VaR"
Sharpe, Sortino, Calmar, max-DD, VaR, CVaR, Cornish-Fisher VaR (fat-tail adjusted), skewness, kurtosis, alpha, beta, win rate, profit factor. 31 unit tests.
šŸ“ packages/be/src/backtest/metrics.ts + metrics.test.ts
āœ“
Verified
"9 anti-hallucination defenses for LLM algorithm generation"
1. Zod schema validation, 2. Primitive whitelist, 3. Cite-or-refuse, 4. Deterministic cross-validation, 5. LLM Jury, 6. 0.85 confidence threshold, 7. Rejection audit log, 8. asOf-frozen context, 9. Adversarial review.
šŸ“ packages/be/src/algorithms/llm/ + docs/12-algorithm-builder.md
āœ“
Verified
"Manipulation detection (spoofing, wash trading, pump & dump)"
7 detection types implemented. Stop-loss hunting, spoofing/layering, wash trading, pump & dump, closing auction manipulation, options flow, real-time spoofing. 14+ DSL primitives. Binance L2 WebSocket integration.
šŸ“ packages/be/src/insights/manipulation-detector.ts + spoofing-detector.ts + stop-hunt-detector.ts
āœ“
Verified
"Cultural calendar seasonality with 30+ events"
Ramadan, Chinese New Year, Diwali, Easter, Christmas, Eid, Thanksgiving, Golden Week, Holi, etc. Lunar calendar calculators (Hijri, Chinese, Hebrew, Hindu). 16 DSL primitives. Seasonality heatmap with t-statistic.
šŸ“ packages/be/src/cultural-calendar/ + docs/83-cultural-calendar-seasonality.md
āœ“
Verified
"Email-to-trading-signals pipeline (mailbox insights)"
Office 365 integration, rule-based processing, LLM entity extraction. 8 DSL primitives (mailbox_sentiment, mailbox_consensus_score, etc.). Timeline heatmap, trending detection, auto-algorithm generation.
šŸ“ packages/be/src/mailbox/insights/ + docs/83-mailbox-monitoring.md
ā—
Partial
"Full Kelly criterion position sizing"
Kelly formula implemented (f* = (pƗb-q)/b) with half-Kelly default. DSL primitives exist. However, win probability (p) is estimated from backtest, not live-updated. Real Kelly requires ongoing calibration.
šŸ“ packages/be/src/algorithms/dsl/evaluator.ts
ā—
Partial
"Regime detection with Markov HMM"
5-state regime classification exists (Bull/Bear/Sideways/Crisis/Recovery). HMM state transitions modeled. However, the Markov model is rule-based approximation, not full probabilistic HMM training.
šŸ“ packages/be/src/fund-manager/regime-strategy-selector.ts
Honesty Note

All "Verified" claims have been sanity-checked against actual code paths. "Partial" claims work but have caveats noted. No "Red" claims are made in current marketing materials ā€” we don't sell what isn't shipping.

šŸ”¬ Quantitative Infrastructure Summary

Component What It Does Technical Detail File Location
Backtest Metrics 13 performance metrics Sharpe, Sortino, Calmar, VaR, CVaR, Cornish-Fisher, skew, kurtosis backtest/metrics.ts
Trade Execution Market impact modeling Almgren-Chriss square-root, bid-ask tiers, slippage simulation trading/services/trade-execution.ts
Tick Classifier Microstructure analysis VPIN, Kyle's Lambda, whale ratio, bot signature, aggressor imbalance insights/tick-classifier.ts
PINN Predictor ML with trading constraints 16 features, 3 heads, Kelly/liquidation in loss function, TensorFlow.js predictions/pinn-predictor.ts
Regime Detection Market state classification 5 states, GARCH volatility, Hurst exponent, VIX quartiles fund-manager/regime-strategy-selector.ts
Kelly Criterion Optimal position sizing f* = (pƗb-q)/b, half-Kelly default, leverage caps algorithms/dsl/evaluator.ts
Risk Guardrails L1-L4 circuit breakers Max DD, daily loss, consecutive losses, event blackout, LLM anomaly algorithms/guardrails/
Pattern Detection Technical pattern matching 40+ candlesticks, DTW templates, Wilson score, regime-weighted shared/src/fvg-detector.ts
Crisis Resilience Defensive stock scoring 5-factor composite, 2008/2020/2022 crisis performance crisis-resilience/
Cultural Calendar Event seasonality 30+ events, lunar calendars, phase analysis, t-statistics cultural-calendar/

šŸ’” Why Would Someone Pay For This?

āœ“ Reasons TO Pay

  • Time savings ā€” Plain English ā†’ algo instead of writing Python
  • Consolidation ā€” One platform vs. TradingView + QuantConnect + data feeds
  • Unique signals ā€” Email intelligence, prediction markets, human-vs-bot
  • Institutional data ā€” 13F/13D filings usually cost $200-500/mo separately
  • Multi-broker ā€” Trade across 5 brokers from one interface

āš  Reasons NOT To Pay

  • New platform ā€” No track record, trust takes time
  • Complexity ā€” 143+ primitives is overwhelming at first
  • Existing tools ā€” Switching cost if already set up elsewhere
  • AI skepticism ā€” "Can AI really trade?" is a valid concern
  • Not autonomous ā€” AI generates algorithms, YOU run them
The Core Value Proposition

The platform's unique value is: "Know when humans are driving the market vs. bots, and trade accordingly."

If the human-vs-bot divergence engine works (correlation between sentiment and price identifying regime), that's genuine alpha. Everything else (AI algo generation, multi-broker, etc.) is convenience on top.