procurement beyond rpa - edgeverve€¦ · procurement beyond rpa: the cognitive future of...

8
Procurement beyond RPA: The cognitive future of procurement in the age of Artificial Intelligence and Machine Learning

Upload: haminh

Post on 20-Aug-2018

216 views

Category:

Documents


0 download

TRANSCRIPT

Procurement beyond RPA: The cognitive future of procurement in the age of Artificial Intelligence and Machine Learning

2 www.edgeverve.com

Catalogs and Requisitions were introduced. Users

could now search for an item in the catalog by providing a text

description, instead of having to know things like Material ID.

Material masters were created, the systems of purchase orders, good receipts and 3-way match

were computerized.

The evolution of procurement technology systems:ERP and MRP Systems: In the first wave, large global organizations invested in ERP and MRP systems to computerize their processes with the intention of getting all information and records in one single source of truth. The primary objectives were process compliance, discipline and data sanctity. Material masters were created, the systems of purchase orders, goods receipts and 3-way match were computerized. However, the primary design considerations for these systems were completeness and accuracy; user convenience was not necessarily a factor.

As system usage increased and the user base became more widespread, user convenience became an important factor.

eProcurement systems eProcurement systems started making a mark with the promise of providing greater convenience to the user than the clunky ERP systems. These systems introduced the concepts of catalogs and requisitions. Users could now search for an item in the catalog by providing a text description, instead of having to know things like Material ID. While this was a significant change from the ERP systems, most eProcurement systems have failed to evolve the user experience in line with the expectations of usability, and remain extremely difficult to use.

For example, the amount of information captured from the users still remains very high. In our experience, we have seen several large global enterprises that continue to have a large number of people deployed in roles such as “preparers”, with the primary job being to prepare Purchase Requests in the eProcurement systems based on information provided by the end-user in a fixed spreadsheet template. Over the last few years, the trend that has been becoming popular is that of Self-service procurement systems.

Traditionally, enterprises have viewed procurement transformation as a linear journey. However, today we are at an inflexion point in the evolution of procurement technology.

www.edgeverve.com 3

Self-service procurement systems

empower the category managers by allowing them to set different channels for

individual commodities and rules for guiding the end-user or a casual buyer towards those

channels.

Self-service Procurement SystemsThese systems go further than traditional eProcurement systems and enable the end-user to raise their own purchase requests, track their own orders, etc. These systems empower the category managers by allowing them to set different channels for individual commodities and rules for guiding the end-user or a casual buyer towards those channels. Further, they empower the casual buyer by guiding them in the buying process across these channels and enabling them to complete the buying process in a manner similar to one that they are used to in their personal lives – search, add to cart, check-out. One of the main characteristics of these systems has been the concept of built-in compliance, ensuring that aspects like preferred suppliers, contracted rates, limits, approval workflows, etc., are all handled behind the scenes without the end-user having to worry about those.

A changing worldWhile businesses in general have been progressing on this linear journey, and specific enterprises are at different stages in their own evolution, today the world is changing. Advances in technologies for automation and Artificial Intelligence (AI) and Machine Learning (ML) are now making it possible for enterprises to take a huge step up and away from where the linear journey would have taken them.

Robotic Process Automation (RPA) is enabling rapid improvements in productivity and efficiency by the automation of repetitive manual tasks. Technology has enabled the redeployment of the people from these repetitive tasks into more value-adding tasks. This has been a driver of both gain in productivity and in several cases, also an improvement in employee satisfaction resulting from redeployment to tasks that are seen as more meaningful and intellectually stimulating.

4 www.edgeverve.com

While RPA is about mimicking

human behavior, Cognitive Automation is about discerning patterns, recognizing context, collating multiple data points

and using all of this for understanding intent and then

behaving in an intelligent manner.

RPA in Procurement:Some commonly automated procurement functions include:

1. the automation of the PR creation in organisations that are using the type of eProcurement systems referred to earlier

2. creating RFX events and sending invitations to the suppliers

3. checking contract compliance during the PR to PO flip

4. bots for budget check

5. bots for invoice entry, and so on

Although RPA is a recent trend and organizations are at different stages in their RPA journey, the real breakthrough is coming via Cognitive Automation.

This is the ability to use Artificial Intelligence and Machine Learning to automate processes.

Today, this is no longer in the realms of science fiction. Cognitive automation is seeing application in various use cases across the spectrum of business scenarios – customer service, ITSM, employee assistance for understanding policies, extracting context and metadata from contracts and so on. In the realm of procurement, cognitive automation is set to fundamentally alter the experience for both the end user and for the CPO’s office.

The cognitive future of procurementCognitive automation brings the ability to combine RPA with intelligence. We see the following as some fundamental shifts in the procurement realm as a result of the emergence of cognitive automation:

Chatbots:With the increase in proportion of Gen Y and millennials in the work force, the expectations from enterprise systems are also changing. Fundamental questions are being asked about the need for eProcurement systems or self-service procurement systems. AI or ML-based chat bots coupled with back-end process orchestration will emerge as a key mechanism for casual buyers to raise purchase requests. Users will be able to ask the chat bot on their mobile phones for the price of an item, swipe through matching options that the chat bot displays, and select the one that has to be bought. The intelligence process orchestration will then take over at the back end, and use multiple mechanisms including chat, email, etc., to drive the process forward for approvals, PR creation, PO creation and transmission, etc.

www.edgeverve.com 5

Spend classification:We see the application of AI and ML making rapid progress in the area of spend classification. Today, ML algorithms are already being used for post-facto classification of spend data. Multiple algorithms with varying levels of mathematical sophistication are being applied to analyze data, find patterns and matches with past classification, draw inferences and classify spend data. We are now seeing the application of AI and ML to near-real time classification of data. Purchase orders will be tagged with the right classification at the time of raising, or Invoices will be classified along with the Invoice entry itself. This will not only improve the accuracy of the spend classification, but also impact the efficiency and timeliness of the analysis.

Spend analysis:Spend analysis is another area where we see chat and AI making an impact. The current practice requires the user to log into to a BI system, choose a workspace, load a data set, apply filters, etc. As opposed to this, we see the emergence of an AI-based chat assistant-driven spend analytics paradigm. With this, the analyst or category manager will be able to ask questions in natural language such as best performing suppliers for a particular category, or top 10 categories by geography by quarter for the last year. The intelligent orchestration will then manage all the data querying, grouping, application of filters in the background and display the charts right in the chat window.

Planning and supply chain management:The other area where we see AI making a significant impact is in relation to planning and supply chain management. AI-based systems will make it possible to combine internal and external data, take inputs from news feeds or market feeds, and use these for varied applications such as generating insights on relative price performance, predicting supply lead time or disruptions, etc.

Artificial Intelligence, Machine Learning and intuitive, natural

language-based chat interfaces are redefining the way the

procurement function operates

6 www.edgeverve.com

How to prepare yourself for this future?Based on our experience, we believe the following are the important ways in which organizations need to prepare for rolling out Cognitive procurement automation.

1. Executive commitment. Introducing automation, AI, chat bots, etc. can cause fundamental changes in the experience for both the end-users and the procurement function, along with a large amount of change for the supplier community. Success here requires sustained effort, perseverance, good execution and governance. Executive commitment right from the very top is one of the most important aspects for all of these to come together and see the initiatives through.

2. IT + business alignment. Clearly, initiatives of this nature are such that neither IT nor the S2P function can drive in isolation. Alignment of important stakeholders from both these functions is therefore key to success. While the objective of automation is to protect existing investments, introduction of AI, chat, etc. brings in new technology where IT needs to be fully involved and take ownership. Similarly, introducing new ways of doing business, changing the way compliance is measured and enforced, etc. require ownership from the S2P function. It is imperative that senior executives of both these functions are completely aligned, as that can drive the alignment of the teams at the execution level.

3. Automation COE. Taking a “Center of Excellence” approach towards a cognitive procurement transformation program can significantly improve the chances of success. A COE has the advantages of pulling in the right skill set from across the organization, empowering this team to take the right decisions and make the right choices, and ensuring a focused effort from a team whose primary goal is the success of the program.

4. Expectations management. Another critical aspect is the management of expectations from the program at all levels. Organizations need to do a good amount of homework to identify the right areas and processes for automation/AI and organizations should not rush into it. Similarly, there are bound to be process changes and changes to the way of working and that need to be communicated well so that the expectations of employees and suppliers are set accordingly. All along the way, during the program, there needs to be continuous reset of expectations in line with the learnings from the program.

www.edgeverve.com 7

5. Clear roadmap with prioritization of “Quick Wins.” Given the varied nature of processes that would be covered in an S2P transformation program of this nature, there is bound to be a mix of use cases where applicability is high and those where automation is not a great business case. For example, a business process area where there is no evident pain point or where the manual effort is too low, might not be a good candidate for automation. Similarly, it could be counterproductive to start the journey with a very futuristic use case that needs too many new investments or large scale changes. It is therefore very important that a clear roadmap is defined. Organisations should start with quick wins – identify the obvious effort intensive areas where automation can quickly show benefits. PR creation and Invoice entry are two good examples. At the same time, organisations should introduce an element of “coolness” that goes a long way in helping the program gain popularity in the organization. One example of doing this is a chat-based help use case. It is relatively simple to achieve this as compared to a full-fledged chat based spend analytics, for example. But giving employees the option to get their purchase related questions on policies, compliance requirements, spend limits, etc. answered via a chat interface can go a long way in increasing the popularity of the program.

6. Measures of success. In line with managing expectations, agreeing to the measures of success is equally important in two ways. First, it helps the entire program team remain focused on the objectives. As with any large enterprise-wide program, there will inevitably be unforeseen obstacles along the way and there will be a need to change course, to reprioritize, etc. Having clearly defined success measures helps make sure that the team always has an eye on the correct goal when they encounter these decision points. Secondly, it also serves as a feedback loop into the expectation management and for ensuring continued executive commitment. Knowledge of what is being targeted and how well the program is making progress towards the target is critical.

7. Choosing the right partner. A program of this nature has to necessarily involve a mix of technology and outsourcing services. Depending on the implementation path chosen, there could be interim stages where outsourced services take up some amount of manual work which then gets automated. Similarly, there could be areas that need a combination of outsourced services and technology to achieve the goal. However, given the very nature of cognitive procurement, most of the areas will focus on technology. In our experience, we have seen several cases where large transformation programs fail because there are too many different organisations involved – too many service providers. To avoid these issues and ensure success of the program, organisations should choose a trusted partner that understands the business and the domain, and that has proven capabilities across technology and services to be able to help manage the entire process end to end.

For more information, contact [email protected]

©2017 EdgeVerve Systems Limited, Bangalore, India. All Rights Reserved. This documentation is the sole property of EdgeVerve S ystems Limited (“EdgeVerve”). EdgeVerve believes the information in this document or page is accurate as of its publication date; such information is subject to change without notice. EdgeVerve acknowledges the proprietary rights of other companies to the trademarks, product names and such other intellectual property rights mentioned in this document. This document is not for general distribution and is meant for use solely by the person or entity that it has been specically issued to and can be used for the sole purpose it is intended to be used for as communicated by EdgeVerve in writing. Except as expressly permitted by EdgeVerve in writing, neither this documentation nor any part of it may be reproduced, stored in a retrieval system, or tra nsmitted in any form or by any means, electronic, mechanical, printing, photocopying, recording or otherwise, without the prior written permission of EdgeVerve and/ or any named intellectual property rights holders under this document. PUBLIC

About EdgeVerveEdgeVerve Systems Limited, a wholly owned subsidiary of Infosys Limited, denes, develops and licenses innovative software products and cloud hosted business platforms. We focus on driving revenue growth, cost eectiveness and protability for global corporations and their business ecosystems across across the world. Visit www.edgeverve.com to know more about our innovations in nancials services, insurance, retail, CPG, life sciences, manufacturing and telecom.

Prashant Vaishnav leads the Procurement Automation practice at EdgeVerve, which is aimed at helping organizations automate their source-to-pay functions through robotics-driven, data-driven and knowledge-driven automation. Prashant has a long experience in helping organizations across both IT- and BPO-led transformations.

About the author:

www.edgeverve.com