Introduction
Enaco is France's leading distance learning platform, offering degree programs from Bachelor's to Master's level. In 2019, 23,000 students graduated through their programs.
Role
From product vision to conversation architecture, dialogue writing, research and testing, I owned the full scope of the project. It was an industry first with no direct competitor.
The assistant lived on the Google Nest Mini, gifted to every new enrollee, and on smartphones, serving 8,000 students across 63 university-level programs.
Process
Before anything else, we needed to know exactly who we were designing for. With data on 23,000 current enrolled users students, we had enough signal to build 3 user personas, each representing a meaningfully different relationship with the school and a different context of use.
But defining who we were designing for was only half the work. We still needed to define what the coach would sound like. So we sat down with 20 students and asked one question: what would your ideal coach sound like as a person? The answer was immediate and consistent. Warm, witty, motivational, expressed through a male voice. Not an institution talking at you. Something that felt like a "coach" who was there for you. That brief became the design constraint for every decision that followed.
User personas based on ENACO student archetype.
Breakdown of the coach, based on 20 surveyed students.
Knowing who and what we were designing for made the scope conversation possible. I mapped every student interaction into 10 intent categories and structured the roadmap into a focused MVP, prioritising the highest-frequency, highest-anxiety questions first, and a V2 that added depth once trust was earned. Scope wasn't just a delivery decision. It was a design decision.
A two-part diagram illustrating the conversation flow with the voice assistant: a global map (left) and a detailed close-up of one branch (right).
From there, we wrote every exchange in SSML, in the Actions Console, the platform for developers to extend the Google Assistant. We were able to control rhythm, pacing, and tone at the sentence level. I treated each dialogue as a micro-screenplay: user intent, assistant response, re-prompt. We worked in tight sprints with the dev team and tested early builds on Nest Mini hardware with volunteers in the academic support team because they knew more than anyone else how a good interaction with students would go. What mattered as much as the words was how the coach handled failure. We applied the personality brief where most designers don't look: error states and fallbacks. When the assistant didn't understand, it apologise but more importantly, it stayed in character.
A series of screens illustrating the process of building the application within the Actions on Google console.
Throughout the build, I worked closely with the Academic Director to define scope, defend priorities, and navigate the tension between institutional thoroughness and product speed. Getting that relationship right was what made everything else possible.
The highlight of this project: seeing the application successfully deployed and ready for users.
Results
600+ daily users, representing 30% of newly enrolled students within 3 months. 95% on smartphones, a finding that immediately reshuffled the V2 roadmap.
First screenshot of a conversation with the ENACO coach on a smartphone.
Second screenshot of a conversation with the ENACO coach on a smartphone.
First illustration created for internal communications
Second illustration created for internal communications
Hindsight
Four months was tight, and so was the strategic brief. The direction was clear from the start: lead with the Google Nest Mini experience. It was the innovation story, the differentiator no competitor was doing. Right call for positioning. But it meant the smartphone experience was treated as an adaptation rather than a surface in its own right.
The result: some responses were too wordy for mobile. A reply that felt natural on a speaker was too slow for someone checking their phone between two meetings.
The smartphone use cases were in the requirements. They just never got the prioritisation they deserved. With more time, I would have pushed for a dedicated mobile sprint earlier, not to challenge the Nest Mini vision, but because 95% of students ended up living in the mobile experience. The signal was always there.