Chemical manufacturing
Powered by
Powered By AWS | Case Study - Manufacturing

$29M EBITDA Gain: How Indorama Ventures Transformed Execution with Agentic AI Workflows on AWS

Deploying AI-driven execution to bridge SAP and the frontline workforce, slashing an 18-week maintenance backlog to zero and reducing contractor spend by 38%.

About the Customer

The customer is a major American multinational that manufactures essential personal care and paper-based consumer products used by millions every day. With a broad network of asset-intensive manufacturing facilities, the organization operates in highly regulated environments where safety, reliability, and execution discipline are non-negotiable. Its operations depend on seamless coordination between maintenance teams, inventory and storeroom managers, inspectors, and frontline operators—working across shifts, sites, and enterprise systems to keep production running smoothly and compliantly.

“Innovapptive helped us bring structure and accountability to frontline execution in a way we simply didn’t have before. By connecting maintenance, inventory, and inspections into one execution layer, we’ve been able to drive measurable improvements in reliability and production efficiency.”

Director of Manufacturing Excellence

Global Chemical Manufacturer

The Problem

  • 18-Week Backlogs: Maintenance queues were out of control due to slow paper data entry.
  • Outdated Operator Round Solution: Manual paper-based rounds failed to capture real-time issues.
  • Hidden "Dark" Data: Critical field observations were lost in paper logs, never reaching SAP.
  • Contractor Over-spend: External labor was needed to plug efficiency gaps.
  • Spare Parts Blind spots: Inaccurate inventory counts led to emergency orders and high carrying costs.

The Innovapptive Solution

Connected Worker Platform

  • Operations Suite: Replaced outdated paper logs with digital shift handovers and rounds, utilizing the AI Detect Agent to capture and categorize real-time issues instantly.
  • Executed AI-Augmented Maintenance: Standardized mobile work execution across crews and implemented the AI Plan Agent to automatically generate work packages, slashing the 18-week backlog to zero.
  • Activated the Spare Parts Control Center: Rolled out real-time mobile cycle counting workflows to systematically eliminate stockouts and sustain 99.5% inventory accuracy.
  • Established an AWS Integration Layer: Seamlessly connected SAP, OSIsoft PI, and OT sources using pre-built, AWS-native integrations via AWS AppSync and Amplify.
  • Launched Value360 Analytics: Deployed live KPI dashboards and automated Digital 5-Why Engines to actively govern contractor spend, reduce MTTR, and track total cost savings.

The Impact

  • $29 Million Realized EBITDA Gain: Direct bottom-line financial impact delivered through digitized frontline operations.
  • $50 Million EBITDA Run rate: Overall reliability, safety, and productivity planned annual savings trajectory at Port Neches.
  • 38% Reduction in Contractor Headcount: Significantly reduced dependency and spend on third-party external labor.
  • 99.5% Inventory Management Accuracy: Total elimination of stockouts and emergency material expediting.
  • Zero Backlog: Maintenance queue fully eliminated, restoring total execution efficiency.
Watch the Story

Empowering the Frontline at Port Neches

See how Indorama Ventures leveraged Innovapptive to digitize their entire frontline operations and maintenance—saving thousands of man-hours, reducing contractor dependency, and delivering hard-dollar EBITDA impact.

Architecture

The Technology Stack Powering 8% OEE Improvement

A comprehensive ecosystem bridging the Physical Plant (OT), the Enterprise System of Record (SAP), and the Innovapptive Execution Layer—powered by scalable AWS infrastructure.

01

Innovapptive Execution Layer

Digitizing the 80% of unstructured field data

Mobile Operator Rounds Shift Handover Intelligent Maintenance Suite Backoffice Control Centres Value360 Dashboards IoT Connectors Powered by AWS AI Agentic Workflows ↓ RACE (No-Code Engine)
02

System of Record & Integration Layer

SAP as the enterprise source of truth

SAP Work Orders · Inventory · Assets
Integration AWS AppSync & Amplify
03

AWS Cloud Infrastructure

Scalable, secure, and governed foundation

Powered by AWS
IngestionAWS MQ · Kafka
ProcessingAWS EKS · Lambda · Amazon Bedrock & AgentCore
StorageS3 · DynamoDB · Industrial Knowledge Graph
SecurityWAF · CloudWatch
04

Physical Plant & OT Layer

Where value is ultimately created

Equipment AssetsPumps · Motors · Valves
IoT SensorsVibration · Temp · Pressure
OT SystemsPLC · SCADA · DCS
AI Agentic  ·  Powered by AWS

Industrial Knowledge + Agentic AI in the Hands of the Frontline

Plant data — SCADA, Historians, MES, ERP, OEM manuals, SOPs — is unified into an Industrial Knowledge Graph and orchestrated by Amazon Bedrock & AgentCore on AWS into 15 role-specific AI Agents that execute real work on the frontline.

15 AI Agents · 5 Domains

Transforming Frontline Workflows

Operations

  • AI DigitizePaper → digital in seconds
  • AI Auto-Fill–75% form time
  • AI AuthorSOPs 5× faster
  • Work Instructions+40% task completion

Maintenance

  • AI Detect40% faster issue detection
  • Troubleshooting–35% diagnosis time
  • AI Plan90% faster planning
  • AI Assist+30% wrench time

Reliability

  • AI 5-WhyRoot cause 3× faster
  • AI Action Rec.–45% repeat failures
  • AI ReferencesKnowledge 80% faster

Safety

  • AI PPE Check–90% violations
  • AI Permits + JHA–70% approval time

Analytics

  • Automated Insights5× faster to action
  • AI Prompt InsightsInstant answers on Reports
Running on AWS

Every agent is grounded in the customer's own SAP, historian, and OT context — orchestrated by Amazon Bedrock & AgentCore across EKS, Lambda, S3, DynamoDB, and the Industrial Knowledge Graph.

See the AI Agents in action →