Generative AI accelerates growth for a global cloud provider

This global company provides essential infrastructure to centralize and secure client data effectively. Recognizing the growing demand for data centers, the company set ambitious plans to expand its business.

One of the key challenges limiting this growth was the availability of material data in a centralized ERP/MDM solution. In its current format, material data was scattered across various documents and templates, making extracting and consolidating information difficult. As a result, teams had to prepare the data and feed it into the BOM manually.

While this manual process was manageable for a few data centers, it was not scalable to support the company's ambitious growth plans. Plus, with an extensive part catalog, even minor errors in the part numbers could lead to costly mistakes, such as ordering the wrong-sized or incorrectly specified components, which then impacted project timelines. Recognizing the need for a more efficient and error-free solution, the company approached Genpact for assistance in building a scalable and flexible system.

We immediately knew generative AI could achieve the speed and scale required for this Herculean task.

Our two main goals were to:

1. Consolidate the data: We wanted to compile the data from all document types into one accessible and understandable format
2. Cleanse and enrich the data using AI: To increase the accuracy of information, we had to guide the AI algorithms and reduce / mitigate false data

The first step was to use advanced machine learning algorithms and natural language processing to extract data from various source documents. Next, we employed optical character recognition (OCR) and machine learning models to extract information from image-based documents. And finally, to enrich the data, we trained a large language model (LLM) using PaLM 2, a next-generation LLM from Google.

The implemented generative AI tool began swiftly collecting and cross-referencing data like never before. What's more, to maintain the integrity of the information, we conducted thorough manual checks using a human-in-the-loop (HIL) approach. These checks validated the LLM models were working as expected resulting in more reliable outputs.

By harnessing the power of generative AI, we improved the part-ordering process. The combination of advanced algorithms, OCR, machine learning models, and manual checks allowed us to achieve the necessary speed and accuracy in processing data, ultimately leading to a successful outcome for our client.

Our gen AI system expedited part-number data collection, down from months to weeks.

Working together with our client's teams, we delivered a future-proof solution that supports:

• Faster speed to market: With generative AI doing the heavy lifting, our client can enjoy a faster turnaround time on data center builds
• Improved decision-making: Clean and enriched data allows our client to plan timelines for data center setup more accurately

• Improved spending and inventory optimization: With reliable and quickly generated material data and the BOM, teams can order the right materials at the right time, enhancing operational efficiency, lowering the total cost of ownership, and optimizing working capital
• Scalability for future demands: The power of gen AI allows the solution to seamlessly adjust to any changes in parts due to supply chains, location requirements, or customer needs

For this company and others in the industry, it's evident that reliable data delivered by gen AI is the future of data centers' expansion.