T3 · ExecutiveCross-Module

Transparent AI — Evidence Chain Review

Not a black box. See exactly WHY the AI recommends what it recommends.
📈 Increase Revenue Improve Efficiency
Directed Acyclic Graph (DAG)Cross-Module Correlation EngineRAG Citation SystemTemporal Fusion Algorithm
Scenario Walkthrough

How It Works

1
Pak Hendra receives an insight notification about cross-module correlation.
2
He opens the evidence chain: a visual causal graph showing rain data → wet fuel report → press energy anomaly, each signal card showing source, timestamp, and confidence contribution.
👤

Pak Hendra

Regional Director (T3)

Glass-Box AI: Traceable Evidence

Black-box AI is dangerous in industrial settings. When NayaCore recommends an action, it provides a transparent, auditable evidence chain. Every conclusion can be traced back to the exact sensor reading, WhatsApp message, or SOP clause that generated it.

📥
10:02 AM

Multi-Source Ingestion

The AI ingests data from disparate sources simultaneously: a weather station (Cakrawala module), a WhatsApp field report (Plantation module), and a power meter (MillOS module).

Real-Time Ingestion Streams
CAKRAWALA: Precipitation 45mm/h
WHATSAPP: "Tandan basah, susah masuk"
MILLOS: Press #2 Torque ↑ 42%
🕸️
10:05 AM

The Causal Graph (DAG)

NayaCore constructs a Directed Acyclic Graph (DAG) connecting the dots. It maps the temporal and spatial relationships between the rain, the wet fuel, and the mechanical strain on the press.

CAUSAL DAG
Heavy Rain
Field Report
↘ ↙
WET FUEL (FFB)
PRESS #2 OVERLOAD
🎯
10:06 AM

Root Cause & Recommendation

Instead of just treating the symptom (the struggling press), the AI identifies the root cause (wet fuel) and recommends a systemic fix based on indexed operational procedures.

Root Cause Analysis
Diagnosis: The mechanical strain on Press #2 is a secondary symptom caused by an influx of waterlogged Fruit Bunches (FFB) due to heavy rainfall at Estate 4.
RECOMMENDED ACTION:Increase Digester #2 retention time by 15% and raise steam pressure by 0.2 bar to pre-dry the fuel before pressing.
CITED: SOP_Processing_v4.pdf (Sec 4.1.2)
10:10 AM

Human Validation

The operator reviews the exact causal chain. Because the evidence is fully transparent, they can confidently verify the AI's logic and authorize the recommendation via WebAuthn.

Operator Validated
Evidence chain verified. Execution authorized.
TX_ID: 0x44f8a...9c21
USER: Operator A (WebAuthn)
ACTION: ADJUST_DIGESTER_PARAMS

📊 Before vs. After

Traditional Operations vs. NayaCore

MetricTraditionalNayaCore
Reasoning ModelBlack-box Neural NetTransparent Causal DAG
Data SourcesSiloed sensors onlySensors + WhatsApp + SOPs
Trust BuildingBlind faith requiredVerifiable evidence chains