ChronoImpact Engine v6.0

Governance-First Strategic Intelligence for Portfolio Decisions

We don't automate decisions. We automate the safety around decisions.

A bounded, governed, latency-aware decision-support engine for financial intelligence.
Not a trading bot. Not autonomous. Not a black box.

Explore v6.0 Full Documentation

Why CIE v6.0

Governance-First

8 structural invariants enforced in code, not policy. Authority never concentrates in AI.

4-Layer Architecture

CIE → PDMEM → RIOS → Human. Each layer has explicit boundaries and authority limits.

Real Formulas

SafeScore, PDBlend, DBScore, RiskScore. Stock-specific calculations, not random scores.

Latency-Aware

ELR + SPD prevent stale signals. Time-decay equations keep prices accurate.

Complete Audit Trail

7-year immutable record. Every decision logged with full reasoning and outcomes.

Regulatory Compliant

NIST RMF, ISO 42001, EU AI Act, Basel, OECD. Production-grade governance.

v6.0 Modelling Premium Specification

JSON1/JSON2 Data Flow

Standardized input/output architecture for integration:

JSON1 Input: regime, signals (W/A/E/S/SAN/C/P/T/L), market data, news

JSON2 Output: stocks with SafeScore, sector impact, portfolio basket, picks

3 Decision Modes

Mode 1 (FAST): ~1 second | Signal-only processing | 70-75% accuracy
Mode 2 (MEDIUM): ~3 seconds | PDMEM calculation + gates | 80-85% accuracy
Mode 3 (SLOW): ~8 seconds | Full CIE+PDMEM+LIL+RIOS | 85-92% accuracy

SafeScore Formula

SafeScore = 0.4×PDBlend + 0.4×DBScore − 0.2×RiskScore + SectorBonus + RegimeBonus

Action Thresholds:

SafeScore > 70: BUY (add to basket)
SafeScore 50-70: HOLD (maintain position)
SafeScore < 50: AVOID (do not buy, consider exit)
Minimum threshold: SafeScore ≥ 20 (eliminates garbage stocks)

Component Calculations

Each stock receives individual scores:

PDBlend = 0.20×P_RSI + 0.20×P_BPP + 0.20×P_EPCP + 0.20×P_SR + 0.20×P_Mom

DBScore = min((rawVolume / marketCap) × 100, 100)

RiskScore = volatility + ((VIX − 20) / 10) × 5

Plus bonuses: SectorBonus (+5 if aligned), RegimeBonus (+10/0/−5)

4-Layer Authority Model

Layer 1 (CIE): Context authority | Geopolitical analysis | Phase classification
Layer 2a (PDMEM): Method authority | Portfolio analysis | Signal generation
Layer 2b (LIL): Zero authority | Quality scoring | Logic verification
Layer 3 (RIOS): Governance authority | Rule enforcement | Audit trail
Layer 4 (Human): Full authority | Final decision | Execution control

3 Backend Microservices

cie-data-server (Port 3003): Market data + news fetching, parallel processing, robust caching
cie-pdmem-server (Port 3005): Portfolio analysis, stock scoring, sector impact calculation
cie-server (Port 3002): SafeScore orchestration, basket generation, final picks

Comprehensive Documentation

11 production-ready documents covering theory, implementation, and deployment

Theory & Glossary

100+ term definitions, 4-layer model, governance framework, all core concepts

View Glossary

v6.0 Specification

Complete SafeScore formula, JSON1/JSON2 flow, 3 decision modes, implementation details

View Docs

Deployment Guide

VPS setup, local testing, monitoring, troubleshooting, production checklist, rollback

View Docs

Interactive Dashboard

Live data visualization, phase probabilities, sector impact, stock analysis

Open Dashboard

Formula Calculators

Interactive SafeScore, PDBlend, DBScore, RiskScore calculation tools

Launch Calculators

Demo & Examples

Real-world JSON1/JSON2 flows, example outputs, scenario walkthroughs

View Demo