Decoding AI’s Transformative Impact on Source-To-Pay and Procure-To-Pay
- AI enhances capabilities of procurement teams, especially in RFP management. Through Large Language Models (LLMs), AI analyzes bids, identifies past collaborators and automates reminders.
- Generative AI goes beyond traditional data analysis by providing contextual insights, enabling actions like email reminders, generating content and offering a nuanced understanding of specific requirements.
- The role of LLMs with unstructured data and the opportunities for P2P automation underscore AI's potential to augment human potential and streamline processes.
September 27, 2023 | Procurement Software 3 minutes read
Artificial intelligence (AI) is making significant strides in reshaping the procurement landscape. Rather than replace humans, it is enhancing their capabilities, forging a new alliance between humans and machines, with procurement teams handling complex tasks and the algorithms doing the mundane.
One of the most striking examples is the power of AI in RFP management. Imagine a scenario where a company is updating its electrical infrastructure—a task that is undoubtedly intricate. "When suppliers submit their bids, a large language model assists sourcing managers by analyzing, comparing, and summarizing," says Rakhi Mullick, vice president of digital transformation at GEP, in a recent webinar (Adopting an AI-First Approach in Procurement: What, Why and How). Such AI tools can pinpoint previously collaborating suppliers, spot inconsistencies in bids and even send automated reminders to those who haven't responded.
But AI's influence isn't one-dimensional.
Over the years, it has evolved, branching into two main types: traditional AI and generative AI.
The former, grounded in deep learning and big data, excels in classifying and harmonizing data. It leans heavily on analyzing past trends to predict the future, especially evident in spend data classification and analytics.
In contrast, generative AI delves deeper. "The spotlight is on large language models that transcend mere data points. These models can interpret contracts, evaluate historical spend and provide personalized recommendations based on a company's preferences,” says Mullick.
One can witness the actual impact of generative AI in various ways:
• Database Queries with Context:
Instead of merely fetching data, AI provides a comprehensive understanding, giving context-rich insights in RFP scenarios.
• Autonomous Agents:
AI doesn't just stop at analysis; it can take proactive steps such as dispatching reminder emails.
• Content Generation:
Need a category-specific RFP template? AI has got you covered.
• Text Validation & Summarization:
With its ability to juxtapose current data with past records, AI offers a condensed view of the procurement cycle.
• Contextual Understanding:
Grasping the nuance of requirements such as ESG certifications becomes a breeze for AI, integrating it seamlessly into processes.
Beyond these, the realm of chatbots is undergoing a revolution. We're all familiar with the rule-bound chatbots of yesteryears. But generative chatbots, Mullick says, can comprehend language, generate content and validate text.
Yet, rolling out such sophisticated bots requires immense preparation, encompassing vast training data, niche expertise and robust security measures.
Several trends are emerging in AI-enhanced procurement.
"The emphasis is now on fostering strategic relationships, automating processes, and shifting procurement teams from administrative chores to strategic initiatives," says Mullick. From refocusing on category management and spend automation to introducing self-service tools where clients can directly input data, the trajectory is progressive.
LLMs have wide ranging benefits, especially when it comes to making sense of data.
Mullick emphasizes on the potential of LLMs with unstructured data.
"In the source-to-contract side, we see more use cases currently. The contracts, the RFPs, the terms and conditions — there's a lot of unstructured data in them, be it in Word, PPT or PDF. That's where clients are resonating with more use cases."
While AI's prowess in dealing with unstructured data is evident, Mullick also addresses its role in the procurement-to-pay (P2P) process. "In P2P," she says, "we're seeing fewer use cases for LLMs. It's more about automation and rule-based AI algorithms than LLMs."
Paul Blake, senior director at GEP, offers a complementary perspective, particularly on the opportunities for automation in P2P. "P2P is ripe for automation because much of the effort is low-value work. Real AI, as in decision support and analysis, is more relevant upstream than in the P2P space,” he says. "AI will take out a lot of the manual human interaction in what is a mechanistic process."
Watch to the full webcast of Adopting an AI-First Approach in Procurement: What, Why and How.
Learn more about GEP’s AI-first approach to procurement transformation.
