French DeepTech Laboratory

Precision Engineering. Computational and Physical Limits.

AEVON Technologies develops advanced optimization and thermodynamic systems for industries operating under extreme computational, economic and physical constraints.

Explore case studies Optimization Platform
LP • QP • SDP optimization
n ≈ 5,000 reference workload
Conservative ×1000 basis vs OSQP + BLAS
Commercial product

Optimization Platform — LP, QP & SDP

Many industrial decisions are not limited by data availability, but by optimization speed. When models take minutes or hours to solve, companies simplify the problem, reduce the number of scenarios, or delay decisions. AEVON is designed to make large optimization workloads interactive.

Reference performance basis

Our conservative public reference is a large QP workload around n ≈ 5,000, benchmarked against a conventional OSQP + BLAS workflow.

×1000
Conservative acceleration basis

Secure deployment model

Clients submit optimization models through a secure API hosted in France and receive optimized results without source-code transfer.

  • French infrastructure
  • GDPR-compliant workflow
  • No client source-code transfer
  • Benchmark on customer workloads
Industry case studies

From batch optimization to operational speed.

The figures below use a conservative n ≈ 5,000, ×1000 reference for large QP workloads. LP and SDP impact is adapted by sector according to the type of optimization typically used. Actual results depend on model structure, density, conditioning and solver configuration.

Hedge Funds & Asset Management

Portfolio construction, factor exposure, risk control and intraday rebalancing.

QP-heavy

The problem

Quant teams often need to optimize thousands of assets under risk, exposure, turnover, liquidity and allocation constraints. When optimization is slow, teams run fewer scenarios, rebalance less frequently, or simplify the model.

What AEVON unlocks

  • More portfolio and risk scenarios during the trading day
  • Faster rebalancing when market conditions change
  • Larger asset universes without turning optimization into an overnight batch
  • Better execution timing for systematic strategies
n≈5,000Reference size
QPMain workload
×1000Conservative basis

Smart Grid & Energy Networks

Battery dispatch, optimal power flow, congestion management and market bidding.

QP / SDP

The problem

Grid operators and energy platforms must constantly optimize under physical network constraints, uncertain renewable production, storage limits, prices and stability requirements. Classical workflows often force operators to reduce the number of scenarios.

What AEVON unlocks

  • More weather, demand and price scenarios before committing decisions
  • Faster battery and storage dispatch
  • Near-real-time support for OPF-like optimization
  • Better integration of renewable generation and grid constraints
n≈4,096Typical case basis
QP/SDPStrong fit
×800+Prudent QP-equivalent

Telecommunications

Routing, capacity planning, spectrum allocation and infrastructure deployment.

LP / QP

The problem

Telecom networks involve large-scale decisions across traffic, routing, capacity, quality of service and investment planning. LP models are common for allocation and flow, while QP appears when smoothness, risk, interference or penalty terms are introduced.

What AEVON unlocks

  • Faster planning cycles for large network models
  • More traffic and failure scenarios
  • Improved resource utilization across dense networks
  • Better transition from static planning to adaptive optimization
n≈8,000Large planning basis
LP/QPCommon formats
×1000+QP potential basis

Logistics & Supply Chain

Routing, inventory allocation, demand scenarios and transport cost optimization.

LP / QP

The problem

Supply chains change continuously: demand, inventory, prices, delivery constraints and disruptions all evolve. When optimization is slow, planners compare only a small number of possible plans and react later to disruptions.

What AEVON unlocks

  • More routing and inventory scenarios in the same time window
  • Faster response to disruptions and demand changes
  • Lower transport and inventory costs through broader plan exploration
  • Ability to move from periodic planning to frequent re-optimization
n≈3,000Typical case basis
LP/QPMain formats
×600+Prudent QP-equivalent
Long-term industrial R&D

Thermodynamic Engine — Decarbonized Energy & Cooling

AEVON Technologies is developing a closed-loop, combustion-free nitrogen cycle designed to simultaneously produce useful energy and industrial cooling for large-scale infrastructure.

Target applications

Data centers, heavy industry and maritime infrastructure require both energy and cooling. AEVON’s thermodynamic program is designed for these dual-demand environments.

Industrial path

Patent filed in 2025. International PCT extension in preparation. The project is structured around an asset-light licensing model with industrial partners.

Founders

Built for research, engineering and industrial deployment.

AEVON Technologies is a French DeepTech laboratory structured around two complementary pillars: a short-cycle software product serving as a financial lever for a long-term industrial R&D project.

Robin Besson

Robin Besson

Co-founder & CTO

A researcher and engineer with a background in applied mathematics and complex systems architecture, Robin works at the intersection of fundamental research, modeling and technological prototyping.

Following studies in mathematics and a degree in architecture at ENSAPVS, he conducted six years of independent research in applied thermodynamics in collaboration with researchers from the ENS. He joined the entrepreneurship program at École Polytechnique and the MATRICE incubator.

At AEVON Technologies, Robin leads R&D and the technological architecture of both pillars: the LP/QP/SDP optimization platform and the combustion-free nitrogen thermodynamic engine.

Valentin Faure

Valentin Faure

Co-founder & CEO

Valentin built his early experience in real estate and later as Sales Director in the sports events industry, where he developed commercial operations, strategic B2B partnerships and complex project execution.

It is on these foundations that he co-founded AEVON Technologies with Robin Besson.

At AEVON Technologies, Valentin oversees strategy, financing, partnerships and the commercial deployment of both activities: the optimization platform and the thermodynamic engine operating-license model.

Benchmark

Test your own optimization workload.

Send us a representative LP, QP or SDP workload. We benchmark it against your current workflow and show the real performance difference on your own constraints.

contact[at]aevontechnologies.com