AR3D

Applied Research · Consulting

Scroll ↓

Senior Applied Research Consultant

Une vision globale, profonde et aguerrie pour les
décisions critiques.

Fifteen years turning hard, cross-disciplinary problems — AI, robotics, 3D vision, distributed systems — into calls that hold up in the real world.

« Impensive » — Impressive & Expensive.

*CGI DE LA VIEILLE ÉCOLE — PAS D'IA A L'EPOQUE

↓

Approach

I work where physics, biology, mathematics, computer science, robotics, 3D vision, AI and distributed systems overlap — and I make the call. Strategy and decisions grounded in what ships, not theory.

How we work together

Half-Day Missions

Fast diagnostic. Decisive results.

Format4 hours, on-site or remote

OutputWritten decision memo within 48 h

PricingFixed fee, agreed upfront

Details +

Fractional Advisory

Your senior voice on every critical decision.

Cadence10–20 h / month, ongoing

ScopeRoadmap, hiring, architecture, risk

PricingMonthly retainer, 3-month minimum

Details +

Strategic Partnership

Equity, success-based, or custom.

StructureEquity, success fee or hybrid

Horizon12 months and beyond

EntryBy application — one or two per year

Details +

Tech Radar

Where technologies stand — my current calls. Click a blip.

—

Selected Engagements

Representative problems. Details under NDA.

Satellite viewing cones projected toward Earth

Space · 3D Simulation Institutional

A satellite that had to exist in simulation before it could exist in orbit.

Move — Built the physically-grounded 3D simulation and photorealistic rendering pipeline behind a satellite digital twin.

Outcome — Work packages accepted and delivered to space-sector institutions.

Earth limb and polar cap — viewing geometry

Ground coverage contours, annotated in dBW

Captures of the visualization software. The visualization layer — viewing cones, footprints, orbits — is my engineering work; satellite models and Earth textures come from the client and third-party astronomy libraries.

Case +

Biomedical · 3D Vision Published

Drug activity on 3D tumor models — with no reliable way to measure it.

Move — GPU-parallelized 3D segmentation with a deep-learning validation loop, running in real time on very high data volumes.

Outcome — New measurement practice adopted; peer-reviewed publication.

Full wireframe reconstruction, tool view

Ellipsoid fit and principal axes — shape metrics

Case +

Product Design · Object Engineering Shipped

A faceted pendant lamp taken from concept to a manufacturable object.

Move — Modeled the facet geometry, resolved structural and lighting constraints, and produced production-ready studio renders in two finishes.

Outcome — A physical object ready for manufacturing sign-off, not just a render.

Independent product-design work, created end to end — no NDA constraints on what's shown here.

Case +

Full context shared in conversation — see References below for institutions and publications.

References

Three tracks of real-world work.

Space · Public Research

Fundamental 3D Research

a new ground for 3D simulation and space research.

▾

Under NDA — no visual cleared for release 3D matter simulation and photorealistic rendering for a major industrial client

Detailed satellite 3D model Satellite digital twin

Painterly rendering research, IRIT Peer-reviewed scientific publication — "Towards Simulation of Handmade Painterly Animation"

Simulation & visualization deliverables Work packages delivered to space-sector institutions

Biomedical Research

3D Tumor Characterization

An approach that introduced new measurement practices in early-stage research.

▾

3D tumour model, volumetric renderHigh-precision 3D characterization of early-stage tumor models

Full wireframe reconstruction, tool viewReal-time 3D reconstruction on very high data volumes

Per-nucleus segmentation outputDeep-learning validation algorithm, GPU multi-parallelized segmentation

Spatial subdivision toolingScientific framework of research-support tools

Quantitative shape metricsAssociated scientific publication — "Evaluation by quantitative image analysis of anticancer drug activity on multicellular spheroids grown in 3D matrix"

PROPRIETARY R&D · INTRAPRENEURSHIP

Internal Framework & Company Track Record

A complete R&D practice built and run as an operating company.

▾

GPU parallelism matrix formulation genetic algorithms internal tooling agentic AI mathematical optimization 3D/GPU framework

R&D — Proprietary Framework

Proprietary 3D/GPU R&D framework
Genetic algorithms
Agentic AI
Parallelism, matrix formulation, mathematical optimization
Complete internal tool suite

Entrepreneurial Leadership — IT Refurbishing

Financial break-even reached
Step change in revenue scale
Commercial development
Established online presence, labels and audits passed
Public network, professional clientele
Reputation built over several years

Co-auteur d’un poster présenté à la conférence NPAR par l’équipe VORTEX de l'IRIT.

Logos reference past professional engagements — not active partnerships or endorsements.

Impact Map

Where a senior call pays off first. Click a point.

—

Where I create leverage

Vingt-deux leviers exemples. quatre familles.

01 / 22

Core Compute

Cost Optimization

Cutting spend without giving up technical performance or strategic capability.

←

→

Contact

Serious inquiries only.

Worldwide · On-site · Remotely

AR3D

Une vision globale, profonde et aguerrie pour lesdécisions critiques.

Une vision globale, profonde et aguerrie pour les
décisions critiques.