The afternoon of the first day at the Ivalua Now event in Paris had a strong AI (artificial intelligence) flavour. (Now there’s an interesting question – if AI had a flavour, what exactly would it be? Fresh mint, maybe? Or aluminium, perhaps?) Anyway, there were also a couple of very good customer cases studies, but we’ll save those for a future discussion and focus on AI here.
Alex Saric, the CMO of Ivalua, started with the context for AI. Procurement leaders are being given more and more priorities, he said, but there are obstacles in their way that stop them delivering strategic decision making and real value. A recent Ivalua study with Forester identified capacity being consumed by operational and manual activities as one barrier, along with a lack of contextual access to insights, and poor data quality as other blockers.
So, is AI the light at the end of the tunnel for overworked procurement functions? There are some obvious applications, such as invoice data capture, data analytics and automating ordering / approval processes, yet, Saric asked, why don’t we have practitioners talking about successful AI case studies yet? A quick poll of his audience supported that, with over 60% of those voting saying they had seen no benefits as yet from AI.
One key to changing that is getting the basis (and basics) right in terms of AI. That means the “data layer” which must provide a “unified data foundation”. That is the data around spend, suppliers, contracts and so on that provides the ammunition or the fuel for AI to work on. If that data is rubbish, then even the best algorithms can’t achieve much. Then the “intelligence layer” (machine learning, NLP etc) and finally the “engagement layer” (how we interact with AI across various procurement processes and tasks) sit on top of that.
If we get that right, then AI can digitize the full P2P process, eliminate manual processes and free capacity for strategic planning – which takes us nicely into the session from Vishal Patel of Ivalua, who told and showed us more about Ivalua’ s approach to AI.
The strategy for the firm is to embed AI onto the core Ivalua platform for multiple use cases across the whole source to pay process. That includes an integrated virtual digital assistant – IVA- to help customers make better, faster decisions. The aims are to support improved decision making for all users by providing easy access to useful insight; to automate more tasks in order to free capacity for more strategic efforts; and to enable a new and different way to engage and interact with solutions.
IVA sits on the platform and interacts with users in natural language. It can;
- Retrieve information on spend, suppliers, etc.
- Help build a shopping cart
- Generate quick insights and display data
- Support users with a voice enabled browser
- Provide user and “how to” type guidance
Users will be able to ask very specific questions – “can you show me our spend per month”? When IVA delivers a chart back to the user, it is then easy to change the axis or the time period or drill down into category spend.
Finally, we had a panel discussion with Sylvie Noël from Covéa, Hélène Paul from Michelin and Aurelien Coquard, Ivalua’ s AI guru, with the ubiquitous Mr. Saric chairing the debate. That again covered AI, although with the title of “Enabling Smarter Procurement”, there were other topics too. Paul talked about the vision in her firm being around creating value from suppliers to benefit Michelin whilst providing a better user experience for everyone involved. Noël talked about value again (“with a capital V”) and working with ecosystems, not just suppliers.
In terms of technology, the panel agreed that this was not just about the systems; “we have to change the way people think, the way we are working, and that includes stakeholders and suppliers”. Soft skills and leadership are key enablers for procurement teams and are as important as technological competence. And AI is not THE solution, but is one of the potential solutions.
Saric asked whether the panel felt a lot of the talk about AI was hype? Noël felt it is real and there are examples in her firm – and she is challenging and pushing Ivalua to come up with more real applications! Paul was perhaps more cautious – “I am not seeing any real application yet, but I do think it is coming, in data, business intelligence, enabling smarter decision making”.
The questions from the audience got into some further interesting issues though. Coquard gave the example of using AI to go through a whole pile of contracts to find risks and issues – a task that would take (wo)man years to do manually. Paul agreed, saying that AI would not replace what humans do but would add intelligence and enable buyers to focus on more added value activities. She preferred the term “augmented intelligence” rather than “artificial”.
Noël is looking to have non valuable tasks automated and pointed out that this could benefit not just the procurement team but the total supply chain. More support in terms of predictive analytics is the dream. But much of the talk had been around transactional or analytical work – did the panel see the scope for AI supporting more strategic work, such as category management and sourcing?
Yes, Noël said, and her firm was beginning to look at new ways (supported by technology) of approaching sourcing and finding suppliers. But we are a long way away from being able to trust the system to make strategic decisions – AI can make suggestions, but human intervention is still needed for decision making (although that might change in the longer term). “Tools can analyse data, but I cannot imagine a tool could build a category strategy, including gaining stakeholder feedback, and properly aligned with the business and the market”.
Another question raised the excellent point that AI may create whole new areas of valuable work for procurement. So (taking Coquard’s point above), if AI does find issues in contracts, that generates additional value-adding tasks, as someone will potentially have to execute tricky negotiations with suppliers. That was a new insight for me – a lightbulb moment really – and one that left me feeling rather more positive than I had earlier. Will AI decimate procurement roles? Or will it enable us to find exciting new routes for procurement to deliver real value to our organisations? Let’s hope it is the latter.
Author: Peter Smith, Managing Director at Procurement Excellence Ltd. Until recently also Managing Editor, Spend Matters UK/Europe.