Implementation Examples – GenAI in Shared Services/GBS
Introduction
During a one-hour Peercast that took place recently, GenAI project leaders from Company A, Company B, and Company C were featured on the Peercast “Real Implementation Examples for GenAI”. The discussion was lively and insightful, with active participation from attendees. (Names and contact details are available to Peeriosity-100 members).
Company Experience: Real Implementation Examples for GenAI
The meeting focused on real examples of implementations for Generative AI in shared services and GBS organizations, featuring panelists from Co A, Co B, and Co C. The discussion explored how companies are implementing AI across enterprise systems, with examples including contract analysis, journal entry processing, and document generation. Key challenges discussed included the need for human oversight to ensure accuracy, managing rapid changes to Microsoft’s Copilot functionality, and addressing governance concerns around citizen developers creating AI-powered solutions. The panelists shared that while AI adoption is growing, with many organizations moving beyond pilot projects to completed implementations, there remains significant work needed around establishing proper guardrails and ensuring data accuracy in AI-generated outputs.
GenAI Implementation Strategies
The panelists discussed their organizations’ approaches to implementing and leveraging Generative AI (GenAI). Rahul at Co A explained that his company has adopted a three-pronged strategy, involving enterprise-led, function-led, and employee-led GenAI practices, with a team of 3-4 people enabling these initiatives. Simon shared that Co B has rebranded its IT organization to focus on data, digital, and AI, implementing a “spoke hub” model with central IT leadership and function-specific teams. Both organizations have enabled Microsoft Copilot and ChatGPT for all employees, with more advanced AI tools like Anthropic’s AI models and Replit AI models available to specific users.
Matt and Galen from Co C discussed their enterprise-wide rollout of Copilot, emphasizing its use as an accelerator and its integration with their RPA team. They highlighted challenges such as rapid changes in GenAI capabilities, security concerns, legal considerations around bias, and the need for governance. Simon from Co B agreed with these points and added that a lack of GenAI knowledge among employees is a significant hurdle. The discussion also touched on the need for training and creating repeatable patterns to ensure responsible AI usage.
All panelists’ GenAI capabilities are being managed by Shared Services/GBS with a scope that includes all areas of the business, not only Shared Services/GBS.
GenAI Implementations
The meeting focused on discussing GenAI implementations and use cases, particularly in contract analysis and accounting. Michael presented poll results showing that many companies were either just starting with GenAI implementations or had limited completed projects, with several companies reporting more than 10 completed projects. The group explored examples of GenAI usage, including Co D’s work on analyzing contracts and extracting data for lease accounting entries. Galen shared that Co C was not currently using GenAI for contract review but was exploring its potential for credit review processes. The discussion highlighted the need for real-world examples of AI implementation to better understand its applications and benefits.
GenAI Integration and Automation Strategies
Simon discussed using GenAI to analyze journal entries, IT tickets, and document templates, highlighting the benefits of rapid development and improved efficiency. Rahul shared examples of GenAI use in contract review, automation of document processing, and product description generation, emphasizing the integration of GenAI with automation for enhanced business processes. Both speakers stressed the importance of human oversight in GenAI-driven processes to prevent errors and ensure accuracy.
GenAI Implementations in Business Operations
The meeting focused on various implementations of GenAI and machine learning across different companies. Galen discussed using LLMs to categorize and assess learning materials for creating targeted courses, while Kirk at Co E shared the development of a chatbot to handle high-volume inquiries in accounts payable. Matt highlighted the use of Copilot for process optimization, and Rahul mentioned the implementation of AgentForce AI software for triaging incoming inquiries and noting ongoing challenges. Alan at Co F inquired about using GenAI tools for automated outreach in accounts receivable, prompting discussions on potential solutions and challenges.
GenAI in Business: Human Augmentation
The panel discussed various applications of GenAI and automation in business processes, focusing on challenges and opportunities. Alan at Co F shared their company’s experience with automating accounts receivable and training processes, highlighting the need for human intervention and personalized support despite automation efforts. Rahul emphasized the importance of teaching employees to build simple GenAI solutions and maintaining human oversight in financial processes to ensure accuracy. The panelists agreed that while GenAI can streamline operations, it should be used as a tool to augment human capabilities rather than replace them entirely.
The call concluded with a comment from Rahul that it might be best to view GenAI as a very smart intern. It has huge potential for greatness, but there is still the need for oversight and review.
Closing Summary
Peeriosity-100 Peercasts feature either a single company or a panel of companies discussing one specific issue or opportunity. Members have complete access to company and participant names with the ability to interact directly with other members. To ensure candid discussions, consultants and vendors are prohibited from membership. For every challenge there are often multiple design options, technologies, and vendors available. Understanding the right approach for your company is a much easier process when you can learn directly from the experience of your peers.
Where have you implemented GenAI? Is your current approach meeting your needs or are their changes that should be considered?
Who are your peers and how are you collaborating with them?
____________________________________________________________________________
Peercasts are private, professionally facilitated webcasts that feature leading member company experiences on specific topics as a catalyst for broader discussion. Access is available exclusively to Peeriosity-100 member company employees, with consultants or vendors prohibited from attending or accessing discussion content. Members can see who is registered to attend in advance, with discussion recordings, supporting polls, and presentation materials online and available whenever convenient for the member. Using Peeriosity-100’s integrated email system, Peer Mail, attendees can easily communicate at any time with other attending peers by selecting them from the list of registered attendees.
Membership is for practitioners only, with no consultants or vendors permitted.
To learn more about Peeriosity-100, please visit www.peeriosity-100.com. Limit of 100 companies.