Investing in artificial intelligence is one thing; extracting measurable value from it is another. At a Tech For Industry Show session, Matthieu Blin, CIO (Motul), joined by Ulysse Duquesne, Manufacturing Lead (Asana), looked back at how his company first structured its project governance with Asana, before experimenting with collaborative AI agents built directly into the tool.

The session opens with two studies. The first, conducted last summer among around a hundred industrial executives from organizations generating more than $1 billion in revenue, shows that two-thirds of them consider AI still at the exploration or targeted-testing stage, and only 2% believe they've fully integrated it across all their operations — even as nearly all of them plan to invest more in AI over the next five years.

The second study, focused on industrial project management, shows that 67% of projects run over their initial deadlines, mainly due to scattered tools and a lack of shared visibility. The session's central message: it's not AI that's the bottleneck, but work organization and management — and that's precisely the issue Asana first helped Motul address, before tackling AI itself.

Motul is a French company specializing in engine lubricants, generating 80% of its revenue from the motorcycle and automotive lubricant segment. Internationalized since the 2000s, the company is now present in around twenty countries and distributes in 160 countries.

The Régénération 2030 transformation plan, launched in 2021 alongside Matthieu Blin's arrival as CIO, is built around four pillars: consolidating the core business (engine lubricants), exploring diversification (notably around electric vehicles) and strengthening R&D (with the recent announcement of a new lab), pursuing an M&A strategy, and developing operational excellence on the IT side.

Matthieu Blin describes a typical initial difficulty for diversified international groups: how do you judge the relative merit of a project investment in South Africa against one in Germany, France or Mexico? Without a cross-functional governance method, IT teams were directly approached by business units in each country, each presenting their request as a priority, with no real arbitration power on the IT department's side.

Rather than deploying a tool reserved for IT, Motul chose to run its POC by involving the communications, marketing and product teams from the start — to verify that the chosen tool, deliberately simple and accessible, met shared needs, without giving the impression of a choice imposed by IT.

The first concrete need identified concerned delivery timeline reliability: rather than committing to an unrealistic schedule, the capacity planning tool deployed makes it possible to identify available resources and projects, and propose a realistic schedule while honoring commitments made. Motul thus moved from a portfolio of more than 100 projects (of which only 30 to 40 actually got completed by year-end) to a more selective approach, with around fifty projects tracked with real visibility, validated upfront by the executive committee.

Unlike other more tightly managed rollouts (tool selection, user testing, training, then deployment), Motul chose the opposite approach: giving teams the tool with a lot of freedom to adopt it on their own terms, supplemented by tutorials and basic training. This approach reached 320 to 350 users, before being reinforced with webinars, best-practice-sharing newsletters, and quarterly presentations by "champion" users.

Deploying Asana let Matthieu Blin demonstrate the value of a PMO function — until then not highly valued at Motul — to the point that the company now has a strategic PMO overseeing all projects, with genuine portfolio control processes.

Asked about the place of AI in his strategy, Matthieu Blin describes a pragmatic approach: raising executive committee awareness more than a year ago, a year dedicated to making IT data reliable and secure, then creating an AI champions unit tasked with identifying business use cases, alongside broad user training. Motul relies mainly on Microsoft Copilot (premium licenses for some users, Copilot Chat for broader upskilling), with an emphasis on training and a "hygiene guide" helping users judge the reliability of the results they get.

The need identified specifically for AI use within Asana concerns tracking projects once they're created: while creating projects in the tool is now well embedded in habits ("no project in Asana, no project at all"), regular follow-up remains insufficient, with statuses sometimes not updated for months. Motul is currently piloting the use of "AI teammates" — AI agents built around specific roles — to automatically generate a genuinely up-to-date status.

The session closes with a demo detailing how a project request can be handled in an almost autonomous way, from intake to execution.

A new project request, submitted through a form, is automatically centralized into a triage project. A "strategic planner" AI agent then analyzes it by cross-referencing its content with all the information available in Asana (strategic goals, related projects): it assesses the project's size, identifies known dependencies, summarizes the need in plain language, checks strategic alignment, and proposes a map of required skills as well as points of caution regarding other ongoing projects. The user can converse with this agent like a colleague, for example asking it to move to the next step.

By cross-referencing required skills with workload and teams' actual availability, the agent proposes detailed staffing, flags skill gaps (for example, no access to a production reference system), and warns of the risk of overloading certain resources. Once staffing is validated by the user, the agent automatically converts the request into a project, integrated into a strategic portfolio for consolidated tracking.

A second agent, dedicated to reporting, is shared with the entire team and works as a genuine collaborator: it has controlled access to certain projects, goals and documents, can generate deliverables (such as presentations), enriches itself from exchanges held on other tools (such as Slack), and builds shared memory over the course of interactions. Asked to prepare an executive committee report, it produces a visual summary of goal achievement, identifies the projects contributing most to identified risks, and proposes governance recommendations.

This session illustrates a fairly representative sequence for successful digital transformation projects: first structure governance and project prioritization with a tool widely adopted by the business, then only afterward layer on agentic AI — capable of speeding up qualification, staffing and reporting for projects — without replacing human decision-making.

Watch the full session, along with the other Tech For Industry Show replays, on our dedicated page.