Maxime Delavergne

Designing and building solutions for professionals and domain experts.

Turning your investments in technologies
 ahem, AI  into actual adoption by users and impact for your business.
If you like that sort of outcome, of course.

01 — How I work

Four principles shape how I work across all projects.

A

Focusing on how it works

I work on how things work — the interactions, the structure, the flows, the seams that decide whether the whole thing holds together — not how it looks.
I build on existing design systems and brands;
my energy goes where it changes the outcome:
the less visible logic underneath. That's usually what separates a solution experts adopt from one they quietly drop.

Engraved pocket-watch movement, exploded view — every gear in its place
B

Experimenting

I treat the experiment as the deliverable.
"Strong hypotheses, loosely held," is my motto.
I build (or prototype) to question and challenge my hypotheses and refine the underlying knowledge that supports them. I build with a purpose, not for the sake of building.
I am experimenting with experiments here: thisisnotready.yt.

Engraved glass laboratory retort on its iron stand — the instrument of trial and error
C

Designing the practice

How we work shapes what we make — the way a solution is designed and built ends up inside the solution itself. So I also design the design: the methods, the collaboration, the systems behind the work. And because you can rarely just apply a method off the shelf, the real work is picking the right one for the job and adapting it to the context.

A hand loom — the frame the work is woven on, and one you keep re-threading
D

Taming AI

AI may be marketed as magic — possibly dark magic — and while it can deliver astounding results, it needs taming more than any other technology to produce what it promises, effectively and efficiently. Forget plug & play (sorry). Building and constantly refining my own system around Claude hasn't just helped me produce new and better work, to the standards I've set — it's also teaching me how to make this technology accessible and useful to people beyond tech roles.

Engraved anatomical brain, lateral view — presented intact
02 — What I do

Five complementary steps.

Five places I work — not a fixed sequence; on a given project we might start anywhere. Each is less a stage than a decision: which problem is worth solving, which users, what to build, whether it landed, and how it holds together with the rest.

JOBS TO BE DONEWORKFLOWSSTAKEHOLDERSBUSINESS
A

Understand

I start by understanding the people I'm designing for: who they are, what they need to get done, how they do it today, and the context around them. I collect qualitative and quantitative data — whichever methods fit the situation.

Seen this in —

Field research platform9 sessions with insights managers and leads across 5 regions — interviews plus real-use-case reviews2023
Perfumery7 perfumers, VP to junior, across 4 countries — prototype sessions, surveys, a diary study2022
Pharma16 multiple-sclerosis patient interviews in Dutch and French; a breast-cancer diary study with patients, nurses and surgeons2021
B

Frame

Research matters when it changes decisions. I use it to prioritise — which users, which tasks, which variables — and to align everyone on clear goals and credible ways to reach them. The framing isn't fixed: it evolves, so we can course-correct.

Seen this in —

Legacy PLMno analytics existed, so usage was inferred from a year of support tickets — the top 20 task types covered ≈ 90% of volume2025–26
Lab servicefrequency data per request type kept correcting screen priorities — and argued the silent requesters back into scope2024–25
Perfumery12 use cases assessed one by one on an impact × time-horizon matrix to steer the roadmap2022
INTERFACE
C

Design & build

Then I design and build the solution — an application, a process, a way of operating. Sometimes on my own, sometimes inside a team; I adapt to the context.

Seen this in —

Legacy PLMa working React proof-of-concept on the dev team's own component library, in about five days2026
Field research platformsketches → clickable prototype → components on Angular Material → specs the dev team built from directly2023–24
Video platforminformation architecture, interaction design, and the studio's first mobile & tablet design system2014–16
UNDERSTANDFRAMEDESIGNEVALUATE
D

Evaluate & iterate

No design gets it right the first time — however good the team. If we don't measure how close we are to the goals we set, we're investing blind. So I evaluate, get feedback, and iterate. That's what turns something shipped into something adopted.

Seen this in —

Field research platforma measured before/after study — mapping accuracy rose from 62.5% to 97.5%2024
Lab serviceiterated with the UI designer until users read the new flow as a clear step up2025
Design practicereintroduced systematic user testing until it became a mandatory step on every project2014–18
PROCESSESPRACTICESSTAKEHOLDERSSTRATEGY
E

Connect

For strategy to reach everyday work — and for what the work learns to reach strategy — an application alone is rarely enough. I also work on the seams: improving processes, improving practices, aligning stakeholders, so the pieces hold together.

Seen this in —

Design practicea UX practice built across enterprise IT — measurement, research ops, analytics, education — argued to leadership twice2023–26
Video platforman active part in the waterfall-to-agile transition between the design studio and development2014–16
Perfumeryan insight repository built for the product's life — it did the onboarding a missed handover couldn't2022
03 — In their words
He is the UX design thinker, the one who needs to understand the context before having any pixels on screen […] Maxime introduced the UX research in the team and created the process attached to it.
Mohamed YahyaouiManaged Maxime at Cisco
He is pragmatic, and perceptive, and is able to balance user concerns, the wider needs of the business, and engineering limitations. A highly skilled, and highly recommended, former colleague.
Paul HardingHead of Product, SocialOptic · Worked with Maxime at Cisco
Maxime is an inspiring team leader, he always tries to improve our processes. He is always the person to count on when production need to be achieved under tight deadlines. HIGHLY recommended!
Raouf BejaouiTechnical Art Director · Allegorithmic
04 — Experience

Fifteen years creating solutions for domain experts
— in tech, fragrance, taste and healthcare.

hvl360°
05 — Education and certifications

With AI, I hear people saying that education and formal training don't matter anymore. I beg to differ. You need people who show commitment to know what they are doing. Formal training says you mean business.

This is Service Design Doing
06 — The other side

This is not ready. Yet

I keep a workshop in the open — applications, concepts, writing... Some of them may be ready. Most of them are likely not. Yet, they are better out there than in my mind or my notebooks.

thisisnotready.yt