I'm Daryna, Content and Research Manager at Mercais. I'm preparing a white paper on actionable AI implementation in procurement, and I'm seeking insights from procurement managers who are actively applying AI or planning to apply it in their processes. Could you share: • A **concrete success story** where AI drove measurable improvements (e.g., cost saving, process speed, risk mitigation), what was achieved, and how?  • A **key challenge or lesson learned** in deploying AI, what didn’t go as planned, and how did you adapt?  • Any **real metrics or outcomes** (e.g., percentage savings, time reduced) or specific use case examples • If you are still in the planning stage, **blockers, fears, uncertainties** you face, and why.

I recently led an AI-driven initiative in our procurement team that focused on automating supplier risk assessments. Previously, the process was manual, taking several days of back-and-forth to verify supplier compliance and financial stability. By integrating an AI-powered tool that cross-referenced supplier data with external risk databases, we cut the assessment time down by nearly 60%. For one large supplier onboarding project, what used to take a full week was completed in under three days. The measurable outcome was not just speed, but also better accuracy—we identified two medium-risk suppliers early on that we might have otherwise missed. That said, the rollout wasn't smooth. One key challenge was staff adoption. Many on the team felt the AI tool was a "black box," and they were hesitant to trust its recommendations. To address this, I organized hands-on workshops where we walked through how the AI categorized risks, and paired it with a manual review stage in the early months. This helped build confidence and eased the transition. Looking back, the biggest lesson I learned is that AI in procurement is as much about change management as it is about technology. Without upfront education and transparency, even the best tool won't gain traction.

As leaders, we introduced AI to forecast raw material costs, aiming for precision and efficiency. The results were clear: 88% accuracy and a 22% cut in budget overruns. Reporting time dropped from days to hours. Yet our first pilot fell short. Data quality issues and unrealistic expectations slowed progress. We shifted strategy; phased data cleanup, realistic milestones, and stronger cross-team alignment. The key takeaway? AI delivers value when leaders champion clean data, set achievable goals, and embrace iterative learning. Waiting for the 'perfect moment' only delays growth. Success comes from testing, adapting, and scaling with measurable outcomes.

A recent project taught me how AI can turn chaos into clarity. We implemented an AI-driven supplier risk model for a mid-sized retail client. Within 90 days, it cut supplier vetting time by 40% and flagged 12 high-risk vendors before contracts were signed. The payoff? A 15% reduction in procurement costs in the first quarter alone. But it wasn't smooth sailing. Data silos slowed us down. Legacy systems spoke different languages, and training staff felt like teaching cats to swim. We fixed this by running pilot programs first; tiny, low-risk tests that gave the team confidence before scaling up. One key lesson: AI isn't plug-and-play. You need clean data, clear objectives, and buy-in from the people who'll use it daily. Start small, prove ROI fast, then expand. That balance of speed and control saved us time, money, and plenty of headaches.