Recruitment · 12 consultants · 5 brands · 250 placements/yearHow Adeis RH powers SEO and recruitment across 5 brands with Dust.
Adeis RH, a 12-person multi-brand recruitment firm, replaced its fragmented AI tools with a unified Dust workspace to automate SEO production, job-description writing and sales-proposal preparation. With Idoption, they built targeted agents that lay the foundations for company-wide adoption.
Adeis RH
Founded in 2004 and based in Nantes, Adeis RH has grown over two decades into a structured multi-brand group. Today, 12 consultants operate under five distinct brands: ADEIS RH (generalist), Human et X RH (social economy), Parcours Public RH (public sector), Gaïa RH (impact and CSR), and Bilan de Compétences.
The firm makes ~250 placements a year, almost exclusively management and supervisory profiles, with consultants dedicated to each sector pipeline. The expertise of a niche firm, the reach of a national player.
From the start, founder Gaëtan Querrec set the tone: stay one step ahead, adopt new tools early.
What was holding them back, before Idoption.
Five brands, five websites, two offices
Running a multi-brand firm with 12 people means wearing a very broad job description. Tristan Girard, head of partnerships for 13 years, drives business development, oversees SEO and marketing visibility across five distinct websites, and contributes to recruitment operations — all at once.
Staying visible on search engines for five brands across five sectors is a continuous, time-consuming effort. Keeping a stream of fresh content flowing doesn't happen on its own.
The ceiling of generalist AI
Adeis RH was no beginner: ChatGPT Pro accounts for two and a half years, internal training. A solid base, but with a visible ceiling — fragile prompting, disconnected from internal data. No way to query connected sources, no persistent agents, no routing of tasks to different models.
We quickly realized it was going to change our day-to-day. We're a small firm; we need to stay one step ahead — at the cutting edge of the tools that save us time.
What we shipped.
Why Dust replaced ChatGPT entirely
In pilot mode, the change was immediate. Connecting Outlook, Notion, Google Drive, Docs, Sheets and Canva to a single workspace, queryable from one interface, changed the day-to-day approach.
Model routing also mattered: some tasks go to Gemini for structure and content architecture, others to Claude for tone. Within a few weeks, users had abandoned ChatGPT.
Dust makes it super easy to connect tools directly. Once you've structured access properly up front, you no longer wonder where the file is or which folder it's in — it just works.
Three agents for three recurring problems
SEO content production: an agent that starts with a keyword benchmark, generates an outline (Gemini for architecture), and produces the draft (Claude for readability). The whole chain in a single space.
SEO ranking tracking: a dedicated agent centralizes tracking across the five sites. A single view.
Job-description writing: an audio brief transcript loaded into Dust alongside client documents, compiled and formatted to the internal standard, ready for review.
Connected to institutional memory
Beyond the three agents, Dust is connected to the client database via Notion. When preparing proposals, Tristan can query past engagements by sector, function or client profile.
It lets me find four or five clients in similar worlds, or in roles we've already handled, to highlight them in a proposal. That kind of search used to take a lot of digging.
What we measured.
Sharper, better-contextualized proposals
The firm doesn't track activity, and in a relationship-driven business time savings don't always map to metrics. But the difference in quality is tangible.
A clear example: a 45-page RFP. Dust made it possible to structure the response point by point, cross-referencing internal documents and brand assets — work that would have taken an entire day.
It lets me be more aligned with client expectations. I can go beyond a generic pitch and show up with relevant context.
100% adoption, among the pilots
The two pilot users completely replaced their previous AI tools. No more hopping between platforms, no more parallel ChatGPT threads. A single workspace, connected to all the data, with agents built around real workflows.
We use it every day. We've abandoned the other tools entirely. And on top of that, it's a French solution. That's worth saying.
What's next for Adeis RH.
From two pilots to a fully AI-powered team
Adeis RH runs on a 12-month ChatGPT enterprise contract for the wider team. When it expires, the direction is set. The pilot phase was intentional: build and validate the agents before any broader rollout.
For mass adoption, the key is to build before deploying, then co-build the use cases with the consultants who'll use them. Quarterly check-ins with Idoption to refine the agent architecture during the rollout across all 12 employees.