Supreme Cortex
Cognitive Precision for Confident Performance
Cognitive Precision for Confident Performance

Supreme Cortex Engineering Analysis Platform

Physics-grounded historian analysis for production support and process engineering teams investigating plant events, performance, and optimization opportunities.

Investor Presentation Pre-Seed
Process flow diagram A process column connected to an overhead exchanger, reflux drum, return stream, and bottoms pump.
The Industry Challenge

The Historian Investigation Bottleneck

The Status Quo

  • Fragmented Evidence: Relevant signals sit inside large exports with site-specific names, units, intervals, and data gaps.
  • Engineering Bandwidth: Process engineers spend hours aligning event windows and reviewing hundreds of trends alongside competing production priorities.
  • Limited Traceability: Reasoning, assumptions, and dispositions are difficult to preserve when each investigation is assembled manually.
Business consequence: slower root-cause work, delayed production decisions, and repeated analysis effort.

The Engineering Analysis Workflow

  • Readiness Screening: Checks time coverage, sampling, gaps, and required process channels before analysis.
  • Confirmed Context: Maps historian tags to canonical roles and units, with engineer confirmation before execution.
  • Physics-Grounded Investigation: Produces hypotheses, evidence, recommended checks, and optimization opportunities for adjudication.
Target outcome: faster, traceable investigations and more confident production-support decisions.
Historian Investigation Workflow

From Export to Engineering Decision

Historian Export Intake

The workflow begins with a CSV or XLSX historian export. The engineer defines the case and event window, then uploads the file for structured review.

CSV or XLSX Historian Export
Case and Event Context
Uploaded Channel Inventory
Controlled Analysis Workspace
Data Readiness Screening

The platform checks timestamps, sampling intervals, gaps, duration, and required process roles before a run proceeds. The readiness summary shows what the data can support.

Time Range & Sample Interval
Data-Gap & Continuity Checks
Required-Role Coverage
Analysis Readiness Summary
Engineer-Confirmed Tag Mapping

Suggested mappings connect client tag names and units to canonical process roles. The engineer reviews and confirms that context before analysis begins.

Client Tag Preservation
Canonical Process Roles
Engineering-Unit Review
Engineer Confirmation Gate
Analysis and Engineer Adjudication

Physics-grounded analysis develops findings and supporting evidence from the confirmed data. The engineer reviews each result, records a disposition, and exports a traceable case report.

Evidence-Linked Findings
Hypotheses & Recommended Checks
Engineer Adjudication
Traceable Case Reports
System Architecture

Historian Export to Adjudicated Case

Historian Export CSV / XLSX Upload Screen & Map Engineer Confirmed Analysis Engine Physics Grounded Findings Engineer Review Case Record Adjudicated Report

Engineer-Governed Workflow

  • Upload and Screen: The engineer uploads a historian export and reviews the readiness summary for coverage, continuity, and required roles.
  • Confirm Context: Suggested tag and unit mappings become analysis inputs after engineer confirmation.
  • Analyze and Adjudicate: Physics-grounded findings, evidence, and optimization opportunities are reviewed and dispositioned in the case record.
Decoupled Prototype Evaluation

Blind Synthetic Validation Protocol

01

Decoupled Data Generator

A separate simulator produces 30 days of physically sound DCS data with injected noise, sensor gaps and randomized tags.

02

Blind Analysis

The analysis engine receives only the design basis and data, with no knowledge of the injected faults, and evaluates the run blind.

03

Separate Automated Scorer

A separate program compares the analysis reports to the hidden ground-truth ledger for accuracy and timing.

Decoupled Simulator Generates 30 days of blind data Analysis Engine Evaluates without fault labels Separate Scorer Computes accuracy & timing Ground Truth DCS logs (faults hidden) Reported detections ?
Blind test: 27 randomized design scenarios · 30 days of continuous 60-second DCS inputs · ground truth withheld from the analysis engine.
Synthetic Proof of Concept

Synthetic Validation Performance

98.2%
Synthetic Blind-Test Accuracy
Correctly identified and categorized fault families across 27 randomized synthetic design runs.
~35m
Median Detection Delta
Detected anomalies at a median of 0.58 hours (~35 minutes) from the start of the fault event.
27 Scenarios
Blind Synthetic Test Coverage
Randomized design scenarios spanning storage, transfer-pump, and air-cooler fault families.
Performance Comparison
Synthetic Blind-Test Accuracy 98.2%
Median Detection Time (From Event Start) ~35 Minutes
Synthetic Design Runs 27
*Tested on 27 distinct design scenarios, representing 30 days of continuous 60s DCS inputs.
Diagnostic Coverage

Target Fault Classifications

Feed Loss Process

Upstream feed failure leading to steady liquid inventory decline and pump cavitation risk.

Details

Tank Overfill Process

Inlet valve stuck open. Inventory rises toward emergency high-high trips or atmospheric release.

Details

Pad Valve Stuck Open Process

Nitrogen supply valve stuck in open position, continuously pressurizing vapor space.

Details

PVRV Stuck Open Process

Vapor pressure safety valve venting gas continuously to atmosphere, causing utility loss.

Details

Padding Gas Loss Process

Upstream N₂ supply failure while drawing liquid, creating structural vessel vacuum risk.

Details

Heater De-energized Process

Electrical immersion heater failure. Fluid cools, risking line freezing or off-spec conditions.

Details

Level Transmitter (LIC) Process

Telemetry level signal freezes. Loop continues static output while inventory drifts blindly.

Details

Pressure Transmitter (PIC) Process

Pneumatic vapor pressure reading freezes. Automated system drifts unchecked.

Details

Pump Bearing Wear Equipment

Progressive bearing degradation across up to 4 pumps. Vibration/temperature trends.

Details
Market Opportunity

A Growing Need for Engineering-Grade Historian Analytics

TAM $3.5B Global historian and industrial analytics opportunity benchmark by 2030 (~10.5% CAGR), within a $50B+ industrial AI software wave.
SAM ~$1B North American opportunity benchmark for process engineering and production support analytics across refining, LNG / midstream, and chemicals, roughly a third of the global benchmark.
Tailwind: Historian analytics and engineering decision support benefit from the broader industrial AI adoption wave (~30% growth benchmark), with North America representing the largest regional share (~37%).
Sources: MarketsandMarkets, Mordor, Fortune Business Insights, EIA and StatCan. Current figures are third-party adjacent-market proxies; dedicated segment sizing is pending customer validation.
Competitive Landscape

Physics-First Engineering Analysis for On-Premises Use

PHYSICS-FIRST
BLACK-BOX ML
CLOUD / SaaS
ON-PREM / AIR-GAPPED
Supreme Cortex
Physics + cloud analytics
Sensor / cloud monitoring
On-prem ML platforms

Engineer-readable physics

First-principles process models provide a traceable technical basis for every finding.

Offline-capable by design

Process analysis runs on premises. Optional signed updates use only a client-approved outbound connection.

Physics-based initial configuration

Synthetic physics models support initial configuration, followed by site data mapping and validation.

Modular unit configurations

Design-basis templates are built to adapt across comparable tanks, pumps, air coolers, and fired heaters.

Go-To-Market

Land, Prove, Expand

01 · LAND

Historian Investigation Pilot

Collaborative investigation of selected historian-export cases with pre-agreed questions, engineering KPIs, and engineer adjudication.

02 · PROVE

Convert to Annual License

Measured pilot performance supports an on-premises annual license and expansion across comparable units at the site.

03 · EXPAND

Expand Unit and Site Coverage

New unit classes and additional sites support stepwise expansion after performance is established.

Founder-led direct sales

20 years of process-operations credibility

Historian and engineering partners

Fits existing data and investigation workflows

Calgary energy-hub density

Direct access to energy operators, EPCs, and service companies

Development Path

Commercialization Roadmap

Phase 1
Historian Analysis Pilots
Structured pilot cases on customer historian exports with pre-agreed KPIs and engineer adjudication of every finding.
Phase 2
Read-Only Historian Connectivity
Scheduled retrieval of approved tags directly from plant historians such as PI, replacing manual exports.
Phase 3
Engineering Case Workflow
Case-based investigation: define the event window, analyze the evidence, record engineering dispositions, export traceable reports.
Phase 4
Multi-Unit Scale-Out
Expanded unit coverage and interconnected flowsheet analysis across plant sections and additional sites.
Team

Built by Operational Experience

Ahmed Sherief

Ahmed Sherief

Founder & CEO

Calgary, Alberta, Canada

"We built Supreme Cortex to help process engineers turn historian data into engineering evidence. It combines operating experience with physics-grounded analysis so teams can investigate events faster, test likely causes, and make production decisions with a traceable technical basis."

20 years in process operations

Global plant commissioning, dynamic start-up, field ops and DCS troubleshooting across refineries, LNG and midstream.

Optimization & debottlenecking

Control-loop troubleshooting and performance optimization grounded in chemical-engineering first principles.

Pre-seed hiring: ML engineer + process engineer to accelerate unit coverage and pilot deployments.
Supreme Cortex. Cognitive Precision for Confident Performance.
Ahmed Sherief

Founder & CEO

info@supreme-cortex.ca

Email

supreme-cortex.ca

Web

Calgary, AB

Canada

© 2026 Supreme Cortex Technologies Inc.

Dual-Parameter Trend: Pump P_101A Deterioration

Tags: P_101A:VI_BEARING.VIB (Vibe) & P_101A:TI_BEARING.TEMP (Temp)

Vibe Current 4.88 mm/s
Temp Current 64.90 °C
5.0 4.0 2.0 0.0 VIB (mm/s) 80°C 60°C 40°C 20°C TEMP (°C) 02:00:00 02:00:10 02:00:20 02:00:30 02:00:40 02:01:00 API 670 VIB LIMIT (4.0 mm/s) Vib Alert (4.88 mm/s) Temp Warning (64.90 °C)
Vibration Amplitude (mm/s) Bearing Temperature (°C)
Illustrative Correlation: High
Finding for Review Asset ID: LP_T101 / P_101A
Detected: 2026-07-02T02:01:00Z

Pump Bearing Degradation (P_101A)

The analysis identified concurrent rises in vibration and bearing temperature on transfer pump P_101A. Vibration amplitude increased from 2.14 mm/s to 4.88 mm/s in a 60-second window, while bearing temperature reached 64.9°C under static speed commands (48.5 Hz).

Vibe Amplitude
4.88 mm/s
Bearing Temp
64.9 °C
Illustrative Confidence
High
Recommended Engineering Checks:
  • Review bearing condition, lubrication history, and recent maintenance records for P_101A.
  • Compare the vibration trend with redundant or local measurements and verify transmitter calibration.
  • Review pump loading and standby availability before selecting a potential response.