In January 2021, as reported by Dave Nyczepir of FedScoop, the General Services Administration’s (GSA) head techie, Chief Information Officer David Shive, announced that he wanted to transform the agency into a “true, end-to-end digital entity.” So far, that’s entailed making electronic copies of the GSA’s intellectual assets—what most people would call “going paperless.” One of the biggest proposed enhancements is the use of artificial intelligence (AI) for predictive analytics, according to Deputy CIO Beth Killoran. Synthesizing big data, drawing more meaningful conclusions for decision makers, and forecasting will be tremendous boons; but given the GSA’s core work, AI could prove to be an equally powerful tool in the federal procurement process.
Digital Transformation: Nice-to-Have Is Now Need-to-Have
Although federal agencies have earned a reputation among critics as lumbering behemoths of bureaucracy and slow momentum, their efforts have consistently inspired developments that have fueled our march toward future breakthroughs. The microchip, GPS, and even the Internet (yes, Al Gore, you helped) are solid examples of innovations pioneered by government agencies and federal contractors, which have taken the mantle of everyday necessities.
In 2017, we saw the passage of the Modernizing Government Technology Act to “allow agencies to invest in modem technology solutions to improve service delivery to the public, secure sensitive systems and data, and save taxpayer dollars.” The trend toward modernization has shifted from primarily military and defense initiatives to intelligence, infrastructure, and even federal procurement.
“GSA recently established shared service product lines for cloud and identity and credentialing services, and has more on the way,” wrote Nyczepir. “The product lines allow for better system integration, which in turn improves security and data exchange, said Beth Killoran, deputy CIO at GSA.”
“GSA aims to procure such AI using its new fourth-generation, enterprisewide Infrastructure Capabilities and IT Operations contract,” he added.
Shive claimed that GSA is already a “heavy” user of AI in predictable places such as cybersecurity. Right now, 51% of GSA workloads have migrated to its cloud, while 22% are handled by managed services off site. The remaining 27% exist on premise. Why shouldn’t procurement also, at some point in time, rank among the “predictable” places that GSA devotes to AI?
Benefits of AI in Procurement
AI is already automating and improving procurement processes for many businesses. It helps expedite time-consuming, repetitive tasks while providing deeper levels of intelligence to help decision makers interpret and act on large, complex sets of data.
Sievo is a global analytics organization. The company published a comprehensive and insightful guide to AI in procurement. The benefits that procurement leaders are already realizing through AI hold obvious benefits to an agency like GSA.
- Spend transparency and categorization. AI’s algorithms can review millions of invoices or related inputs and then classify them into appropriate categories and subcategories through machine learning.
- Vendor matching. With the wealthy of supplier data contained in the GSA database (e.g., category codes, pricing data, purchase order history, NAICS, socioeconomic status, and more), AI can automate the vendor matching process, making recommendations that streamline the selection of best-suited vendors, including smaller firms that could otherwise go unnoticed.
- Anomaly detection. Machine learning algorithms can detect and reveal atypical or anomalous circumstances, such as unexpected changes in purchase prices for a commodity or from a specific supplier.
- Capture supplier or market data. Through natural language processing, the AI can look for and capture data on suppliers or specific markets. “For example, tracking social media channels for signals about suppliers’ risk positions,” Sievo noted. To GSA, there could be wider applications here. One of GSA’s interests in onboarding vendors is understanding how they will market their products or services. AI could provide a much clearer picture of the marketing channels being used, the effectiveness of those efforts, social networking strategies, and more.
AI Can Optimize GSA Procurement Processes
In her article for SourceToday, Bridget McCrea covered an AI research study specific to procurement, conducted by Ivalua and Forrester Consulting. Several of the findings further demonstrate the positive impact that a more encompassing adoption of AI and machine learning could hold for GSA.
Smarter Procurement Process
“In terms of the practical uses of AI,” McCrea explained, “Ivalua’s study found that AI may have the most impact in alerting the enterprise and suppliers to supply chain disruption (44%), recognizing and flagging supplier compliance issues (39%), and quickly identifying instances of fraud (37%).”
Smarter Invoicing and PO Processes
The increasingly sprawling and complex datasets that procurement teams receive on a daily basis create administrative burdens that jeopardize their control and management of the data. “Two of the biggest areas flagged by respondents as having the greatest potential for automation is invoice processing (51%) and approval of proposed purchases (35%),” Ivalua reported.
GSA could leverage AI to automatically review and approve purchase orders. Machine learning is being used in accounts payable automation, which could help GSA speed up payment workflows and detect fraud, if deployed.
An Investment in Ongoing Innovation
Artificial intelligence and machine learning aren’t simple programs that an agency can install and then let the digital magic unfold on its own. Data will need to be fed into the system and supervised by human administrators who must also help teach the platforms what to look for and do. But the applications already being explored by commercial procurement leaders point to powerful advantages that could reinvent how federal procurement processes evolve.