How a conglomerate is building an AI-ready foundation with engineering content

This case study highlights how a leading South Asian conglomerate strengthened asset reliability and operational resilience by deploying OpenText Content Management (Extended ECM for Engineering Content) integrated with Oracle E‑Business Suite.

With more than USD 1 billion in annual revenue and 10,000 employees, the company runs three large‑scale plants 24×7 to sustain agricultural production across the region. Unplanned downtime can cost over USD 100,000 in just two hours, making real‑time access to technical documentation and maintenance records business‑critical.

Recognising that engineering teams were constrained by fragmented data across departmental systems and local storage, the organisation conducted a transformation that demonstrates how industrial manufacturers can use enterprise content management to maximise uptime, empower field engineers with accurate insights, and prepare for the next phase of operational AI adoption.

A leading South Asian industrial conglomerate with more than USD1 billion in revenue, around 10,000 employees, and a fertiliser division serving roughly two million farmers, manages three large-scale plants operating 24×7. Ensuring continuous availability of critical manufacturing assets is key to keeping production lines running and protecting the livelihoods of millions in the broader agricultural supply chain.

Problem statement

Unplanned downtime at the fertiliser plants could cost the business over USD100,000 in as little as two hours, making asset reliability a strategic priority. Asset reliability is tightly linked to engineers having push-button access to the right maintenance manuals, technical drawings, and repair histories. As their director of mobility and digitisation explained, “If we have to shut down a production line, the cost impact is immediate.”

In the past, engineers relied on technical documentation and asset records scattered across Oracle E-Business Suite, email inboxes, shared drives, and employee laptops, creating information silos that slowed decision-making, increased compliance risk, and made it harder to maintain assets and avoid unplanned outages.

Solution: building an AI-ready content layer

To modernise asset information management and prepare for AI, the conglomerate selected OpenText Content Management (Extended ECM for Engineering Content) as a secure, centralised enterprise content platform tightly integrated with Oracle E-Business Suite. “We aimed to enhance our approach to asset management by centralising all our records in a single, secure enterprise content management platform,” noted the director, underscoring the strategic move away from fragmented repositories.

By tagging each record in OpenText Content Management with the corresponding asset number in Oracle, engineers can see all documentation, repair history, and related correspondence for an asset in one place, enabling faster troubleshooting and more reliable maintenance decisions while creating a governed, AI-ready content layer for future analytics and generative AI use cases.

Implementation: consolidating data for AI at scale

Working with OpenText and a certified partner, the company migrated more than three terabytes of asset management data, consolidating thousands of records from Oracle, email, employee devices, and other repositories into OpenText Content Management. The director highlighted the train-the-trainer model that equipped representatives from each engineering team with know-how as a key success factor. Clear guidance at every stage and open lines of communication ensured the resulting repository was complete, consistent, and structured to support AI-driven retrieval and automation going forward.

Measurable outcomes: reducing time-to-information to less than 5 minutes

Today, more than 500 engineers rely on OpenText Content Management as their primary asset information hub, typically finding the documentation they need in less than five minutes — around 70 percent faster than before. “Our engineers can make better-informed decisions, faster. On average, our teams can find what they need in OpenText Content Management in less than five minutes — 70-percent quicker than before,” said the director. Moving records off the transactional Oracle platform into OpenText has delivered a 15 percent improvement in end-user response times for the asset management application and a 40-percent data deduplication rate, improving compliance and creating a curated content layer that is ideal for AI.

How the AI works behind the scenes

OpenText Content Aviator sits on top of the centralised OpenText Content Management repository and uses retrieval‑augmented generative AI to find and summarise information. When an engineer asks a question in natural language, the system interprets the intent, semantically searches governed maintenance manuals, repair histories, and asset records (respecting access controls), and retrieves the most relevant passages.

These snippets are then passed to a large language model, which generates a grounded answer with links back to the original documents, collapsing search, reading, and cross‑checking into a single interaction that typically takes less than five minutes. OpenText says OpenText Aviator is different from regular AI; because of the business context it is tied to, it has a full understanding of real business processes and generates responses in context.

Future roadmap: operational AI for engineers

Looking ahead, the company plans to use OpenText generative AI capabilities so engineers can find asset data even faster using natural language, further reducing time-to-information and enhancing decision quality.