How Applium reached 90% AI adoption in 29 days with Dust
- Industries
- ConsultingTechnology
- Company Size
- 51-200
- Department
- Company-Wide AI Adoption
Key Highlights
- 90% of users onboarded within the first 29 days of deployment
- 20 min to 1 hour saved per meeting with automated meeting minutes
- 70% of collaborators equipped with personalized AI agents
- Hours to days saved per tender with an AI-powered pre-sales agent
This story was written by Idoption, a Dust Platinum Partner with strong expertise in the SAP and IT services industry. It shows how Applium, an SAP specialist consulting firm, deployed Dust to augment its consultants’ impact and accelerate operational performance. With Idoption's support, Applium went from zero to 90% AI adoption in under a month, building personalized AI agents for 70% of its team and integrating Dust directly into its SAP ecosystem.
About Applium
Founded in Paris in 2002, Applium was created with a clear ambition: to deliver SAP services that combine deep expertise, practical innovation, and long-term value for clients. Built by a team of ERP specialists with strong field experience, the company has grown steadily by focusing on complex transformation projects where business understanding and technical excellence are equally critical.
Beyond its expertise, Applium is also recognized for the strong sense of belonging shared by its employees and for a culture rooted in genuine human values: listening, knowledge-sharing, trust, and care. This people-centric approach has shaped the company’s development from the beginning, fostering long-term relationships both within its teams and with its clients.
Today, Applium is a 150-person firm with ten agencies across France, a subsidiary in Portugal, and €22 million in revenue. In 2024 and 2025, the company was recognized as SAP Partner of the Year — a milestone that reflects both its growth and the depth of its expertise. With an average of 16 years of SAP experience per consultant and 80 client references spanning sectors from aerospace (Dassault Aviation, Daher, Sabena Technics…), Industry (Ascodunes, PAM Building, Verescence…) to retail (Boulanger, Cultura, Maison Johanès Boubée/Carrefour…) and the public sector (Adoma, Business France, Etablissement Français du Sang, Radio France…), Applium has built its reputation on one thing: expertise.
That expertise is also what makes AI adoption at Applium particularly significant.
Challenge: When technical excellence is no longer a sufficient edge
The imperative to stay relevant
In the SAP services sector, competition is relentless. Clients, major industrial groups, large retailers, public institutions, expect partners who are technically sharp, fast, and always innovating. Applium decided to get ahead of the curve.
"We are firmly convinced generative AI is a non-negotiable to stay competitive in our sector. We wanted to remain a reliable, relevant, and high-performing partner for our clients — and stay attractive to our own consultants."
— Joseph Idiartegaray, Director of the Support & Maintenance Center and AI lead, Applium
Joseph Idiartegaray brings 20 years of SAP experience to his role leading Applium's Support & Maintenance Center Alongside Jérôme Basselin, Director of Operations, he also oversees cross-functional initiatives (including the company's AI transformation).
But before moving, two prerequisites had to be met: data security and governance. As a firm managing sensitive SAP environments for our clients, Applium couldn't move forward
without ironclad answers on data security and governance.
Dust provides enterprise-grade security guarantees — including EU-based hosting, zero data retention by third-party model providers, SOC 2 Type II certification, full GDPR compliance, and end-to-end AES-256 encryption — ensuring that Applium's client data is never stored, exposed, or used to train AI models.
The bottlenecks eating away at consultant bandwidth
Beyond the strategic imperative, concrete operational friction was slowing the team down every day:
Pre-sales & tenders: Processing complex RFPs meant reading large volumes of documents, coordinating go/no-go decisions, and managing writing efforts across contributors. Each opportunity evaluated consumed hours — sometimes full days — for a team whose core value is expertise, not administration.
Meeting follow-up: Writing meeting minutes was inconsistent at best, skipped altogether when deadlines pressed. Knowledge from client discussions simply disappeared.
TMA ticket management: Applium launched its TMA offer back in 2010, and it has become a cornerstone of the business. But identifying recurring patterns across hundreds of support tickets still required manual effort from account managers, making it hard to surface systemic issues quickly and proactively advise clients.
Solution: A POC built to deploy, not to demo
Why Dust won
Applium came to Dust through a direct recommendation via Idoption's network.
But the decision required rigor. The team needed clear answers on three things: data security, GDPR compliance, and the ability to work independently of any single LLM provider.
Dust addressed all three. Once satisfied, Applium moved fast.
"Once Dust addressed our security, GDPR, and data governance concerns, we took an empirical approach. We wanted a powerful tool, independent of any single LLM, that could connect to our ecosystem. We tried it — and the gains are phenomenal while the limits are very modest in comparison." — Joseph Idiartegaray
One month, five use cases, all in production
Applium followed a focused approach co-designed with Idoption.
Core team: 5 champions* + ~15 people in an extended group
Timeline: 1-month pilot, real use cases from day one
Scope: 4-5 targeted use cases, each built to go directly into production
The key principle: nothing built during the pilot would be discarded. Every agent developed, every workflow tested was shipped directly into the final production environment. Zero wasted cycles.
*A Champion in a Dust AI POC is a key user within the client organization who actively promotes, tests, and drives the adoption of the Dust AI solution among their colleagues, acting as the main internal advocate and point of contact between the end users and the project team.
Training: themed sessions over open desks
Post-pilot, Joseph took charge of internal adoption. He first tried open-desk drop-in sessions — useful for the opening week, then quickly losing traction. What actually worked: focused thematic sessions on specific topics — creating personal agents, building automation triggers, managing emails with AI.
As a Dust admin, he also used the platform's usage statistics directly: identifying who used which features, targeting less active users for support, and turning power users into peer advocates.
The agents that went into production
Three use cases moved immediately from pilot to daily operations:
Pre-sales & RFP agent: Helps the team evaluate go/no-go decisions on tenders, process large procurement documents, and coordinate writing. It directly targets one of Applium's most time-sensitive workflows — winning the right contracts against well-resourced competitors.
Personalized AI agents: Each consultant received a personal AI agent configured for their specific context. Today, 70% of the company is equipped. The impact, as Joseph puts it, is "impossible to measure, but considerable."
Automated meeting minutes: Built on Dust's Teams transcript integration (released April 14, 2026), this agent generates structured summaries automatically from meeting recordings. Simple, immediate, and used consistently where previously nothing was being written at all.
A fourth use case — aAutomated analysis of TMA support tickets — is currently running as an internal pilote, with plans to turn it into a client-facing product.
Results: 90% onboarded, measurable gains, in 29 days
Adoption numbers that speak for themselves
Applium deployed Dust at the start of April 2026. Twenty-nine days later:
- 97 out of 120 users connected and registered (~90%)
- 60% are regular, recurring users
- 30-40% are heavy users who rely on Dust for core daily work
For a company of seasoned SAP specialists (not early adopters by nature), these numbers are remarkable.
"We don't need to prove the wall exists. The ROI is obvious. We'd rather focus on adoption and making sure everyone truly becomes an augmented consultant." — Joseph Idiartegaray
Where the time savings show up
Meeting minutes: 20 min to 1 hour saved per meeting — with the added benefit of systematic production of notes that simply weren't being written before
Tenders & RFPs: A few hours to a few days saved per opportunity, enabling the team to evaluate a wider pipeline with greater precision
Personal agents: Deployed across the company for writing, summarizing, and researching — scope too broad to quantify individually, but pervasive in daily work
Three populations, one direction
The transformation created three distinct user groups within the company:
Power users (10-20%): Dust is the backbone of their work. Every significant topic involves AI.
"There is not a single major topic I work on without AI today. It's a genuine accelerator and a drastic change in how I work." — Joseph Idiartegaray
Regular users (50%+): Use Dust as a strong daily tool — not omnipresent, but a consistent accelerator for specific tasks, much as they previously used ChatGPT or Mistral, but now within a governed, connected environment.
Users still ramping up: A small minority, actively supported through targeted sessions and peer encouragement.
An ethical, deliberate approach
Worth noting: Applium doesn't advocate for using AI everywhere indiscriminately. Joseph is explicit about choosing the right model for the right task, maintaining responsible consumption, and ensuring AI augments — rather than replaces — human judgment. For a firm whose value proposition rests on expert insight, that distinction matters.
What's next: SAP integration and becoming a client-facing AI partner
Deepening the ecosystem
The next frontier is technical depth. Applium has already connected Dust to its internal SAP system — enabling consultants to work directly inside SAP with AI assistance, run table lookups, and handle technical queries through natural language. For a firm where 16 years of average SAP experience per consultant is a selling point, putting AI directly inside the systems their experts know best is a natural evolution.
The next question: how to extend this securely to clients. Data security, RGPD, and governance requirements need to be fully mapped before a client-facing SAP-Dust integration becomes viable — but the roadmap is clear.
Beyond SAP, the team is building out automation infrastructure with N8N, positioning Dust as a central — but not exclusive — actor in a broader AI and automation architecture.
A new product line: AI-powered Level 0 support
The TMA ticket analysis pilot has a commercial application in view. The vision: a subscription-based offer where clients receive an AI-powered first analysis of their support tickets as a Level 0 response layer, with escalation to human consultants for Level 2 and 3 issues. For Applium's clients — industrial groups, retailers, public institutions running complex SAP environments — this could mean faster response times at lower cost, with Applium's expertise available on demand. A new business line, born directly from internal experimentation.
Becoming an AI partner for clients
Applium is also pursuing formal partnership status with Dust — enabled through Idoption's network. The ambition is twofold: deliver better services to existing clients through AI-enhanced consulting, and become a recognized AI partner that guides clients through their own transformation — for SAP-specific use cases and beyond. Thirty new hires are planned for 2026. AI won't be an add-on for them — it will be part of the job from day one.
"One month after launch, it's quite phenomenal to be where we are. Honestly — it happened fast." — Joseph Idiartegaray
Key takeaways for SAP and IT services firms
Build your pilot to survive, not to test
Don't run a pilot in a sandbox. Build it with real use cases, real users, and a commitment to ship everything directly into production. Applium lost zero work during the transition from pilot to deployment.
Champions over mandates
A small group of motivated power users will drive more adoption than any company-wide directive. Identify them early, equip them well, and let them lead from the front.
Solve security first, then move with conviction
For IT services firms managing client data across regulated sectors, governance is non-negotiable. Resolve it upfront — then accelerate. Once Applium was satisfied with Dust's security posture, they moved decisively.
"Stop hesitating and go for it. If you hesitate, you'll miss the train. We had no certainty about everything — we tried, and the gains are phenomenal. The limits are very modest compared to what you get." — Joseph Idiartegaray
Joseph also flags a broader challenge his industry will need to address: "In a few months, someone who hasn't been helped by AI will genuinely be non-competitive. AI in the hands of experts is fabulous — but we'll also need to figure out how to train the experts of tomorrow when beginners are almost replaced by AI." For a firm that has spent 24 years building a culture of expertise, that's the next question worth taking seriously.
Interested in learning more about how Dust can help your team? Visit our solutions page or reach out to our sales team.
