AM/NS India Cuts Vendor Invoice Processing Time to One Sixth

Steel major, AM/NS India implements UiPath Document Understanding to streamline vendor invoice processing, ensuring timely vendor payments and GST reconciliations.

ArcelorMittal Nippon Steel India (AM/NS India), a joint venture between ArcelorMittal and Nippon Steel, is India’s fourth largest steel producer. Established in December 2019, within three years it has grown to become one of India’s leading integrated steel producers with Crude steel capacity 9 MTPA.

Ramping up operations parallelly in the most cost-efficient manner has made this unprecedented growth possible within a short time span, and UiPath has enabled the company to grow exponentially in a cost effective manner.

The scale and scope of its operations means a huge portfolio of vendors that are critical to maintaining the supply funnel. It currently works with 10.5K+ vendors, processing on an average 300,000 invoices per annum. Timely and accurate processing of vendor invoices and payments is what helps the company maintain a healthy cash flow and meet contractual obligations. Rohit Sharma, Head Working Capital, AM/NS India credits this to automation.

We realized early on in our journey that automation is a key lever in supporting our growth vision. Despite the exponential rise in volumes, Automation has helped us to maintain strict deadlines and improve processes in a cost efficient manner”

Rohit Sharma, Head Working Capital, AM/NS India

Challenges in Meeting Vendor Payment Timelines

Coming to this realisation itself was a journey for Sharma and his team. Prior to automation, AM/NS India had unstructured and decentralised invoice processing methods across various locations. Even with a large team of 60+ running the process, the turnaround time (TAT) for vendor payments was high causing delays besides low accuracy levels. The TAT for invoice processing was around 7-10 days with less than 40% vendor payments were on time.

As volumes increase so does the complexity. Manually processing different invoice formats – physical and digital – with a mix of both structured and unstructured data is highly resource intensive and prone to errors. Some vendors also submit invoices in formats different from the standard norms, compounding the complexity further. Meanwhile, certain tasks such as invoice authenticity checks across 25 Goods and Services Tax (GST) rules can be particularly challenging when done manually.

Convinced on Automation

As the finance function set out to address the challenge, it was clear in its goal of a long-term solution that will meet the future scaling up requirements. Finding Robotic Process Automation (RPA) to be the right fit, the team undertook an extensive evaluation process. It’s strong Document Understanding and Cloud Orchestrator features gave UiPath an edge. It’s strong API integration capabilities and easy interfacing with the internal AM/NS India systems made the UiPath solution further suitable to meet the company’s requirements.

We wanted a technology which could help us sync with the robust way of digital process automation within AM/NS India.

Rajesh Chitti, Lead RPA - Information Technology, AM/NS India

Process Improvement Hand-In-Hand with Implementation

UiPath’s collaborative agile program management under the leadership of AM/NS India’s Chief Financial Officer Amit Harlalka, and AM/NS India's Head Working Capital, Rohit Sharma, combined with the strong project management skills brought in by AM/NS India's Information Technology, RPA Lead, Rajesh Chitti, helped ensure fast delivery.

Starting with automating the material invoices at its main location, the company onboarded other locations with the creation of Accounts Payable Centre of Excellence at Hazira. The implementation was completed with the setting up of the Vendor Invoice Portal for invoice submission and service invoices.

Making the transition from manual process to automation, AM/NS India saw an apt opportunity to redesign the process to make it lean and standardised.

Lean process principle was adopted to eliminate the redundant and non-value steps within the process, and aligning with internal teams was done for process re-structuring.

Rajesh Chitti, Lead RPA - Information Technology, AM/NS India

Invoice Processing Time Down by One Sixth

Post implementation the complete end-to-end vendor invoice processing from the time of invoice receipt to payment processing is managed by five bots from UiPath - two AI bots and three regular bots. Today, these bots process over 300,000 invoices annually with over 90% accuracy.

Automation has also helped the company keep up with the growing volumes with minimal cost addition. While the volume of invoices has gone up by 150% since 2021, the headcount addition was only 5%.

With each invoice taking 1 day to process post automation as opposed to 7-10 days earlier, AM/NS India has saved significant time on each invoice. This has not only ensured timely payments to vendors but also freed up 15-20 manpower resources from mundane repetitive tasks. The freed resources have been further trained and reallocated to higher value tasks, improving internal resource utilisation as well as employee satisfaction and retention.

Document Understanding Makes Data Extraction and Validation Easy

The intelligent automation approach with Document Understanding is core to the value delivered by UiPath in transforming the vendor invoice process. AM/NS India is using the auto validation process as ML and Document Understanding with AI feature. Document Understanding is an AI image reading technology that combines AI and RPA for end-to-end document processing.

These intelligent automation features have been particularly impactful and helping us enhance the capabilities and value beyond the regular automation of activities, which is more of a rule-based approach.

Rohit Sharma, Head Working Capital, AM/NS India

Starting with capturing data from invoices, the bot picks the invoice from the in-house invoice portal where vendors upload their digital invoices (both native and non-native). The Intelligent Form Extractor component in Document Understanding then extracts different fields from the invoice and validates the action. It can read and extract data from different types of invoices - digital, scanned and e-invoice – using AI.

A structured validation process follows with the bot step by step checking for invoice authentication, cross-checking all 25 GST rules, and validating PO, material/service items and value match. Thereafter, the bot checks for GRN/ service entry and quality check before finally processing the invoice for posting in SAP.

Sharma informs that a structured approach to invoice validation and processing has helped standardise key components of vendors invoicing, including PO for pricing, invoice data for pricing as per PO, single GRN for single invoice as a practice, GRN fields in SAP for bot processing, invoice patterns, and taxation rules and control checks among others.

The UiPath cloud orchestrator provides centralised visibility, management and real-time tracking of each invoice during the end-to-end process. Thus, removing people dependency for information and timely response to vendor queries.

Future is Intelligent Automation

The accounts payable success story has given confidence to AM/NS India which has extended automation to other areas, including enterprise use cases in logistics, HR, operations and marketing. As a testimony of its commitment to automation, the company has set up an RPA CoE along with citizen development program, enabling end users with business specific automation at individual or department level, especially in the manufacturing division.

By end of 2024, we are targeting to cover all business units with enterprise automation and create a strong workforce of citizen developers to accelerate the automation skillset at AM/NS India.

Rohit Sharma, Head Working Capital, AM/NS India

The larger vision of AM/NS India is built around intelligent automation as a key component of its overall business transformation agenda.

““Intelligent Automation is the next move to eliminate dependency issues on exceptions. AI/ML and analytics related solutions can be easily achieved with intelligent automation solving timebound business problems across both new and legacy technologies. And we are keeping an eye on the technology developments and innovations at UiPath in this area.”

Rohit Sharma, Head Working Capital, AM/NS India