Hybrid Quantum-Classical AI · Gatineau, QC

Molecular discovery,
accelerated by
quantum AI.

Quimiq builds hybrid quantum-classical generative models that collapse years of laboratory iteration into weeks of computational screening — identifying novel molecules and CO₂ capture materials at unprecedented speed.

Request Early Access See how it works
100×
faster than lab screening
12k+
candidates per run
4
technology layers
Top candidate
QED 0.94 · SA 0.81
Screening queue
12,480 structures
Hybrid Quantum-Classical AI Sorbent Material Discovery Q-VAE · Q-GAN · Quantum Diffusion CO₂ Capture Optimization NISQ Hardware Validation DAC System Design Gatineau · Ottawa · uOttawa Hybrid Quantum-Classical AI Sorbent Material Discovery Q-VAE · Q-GAN · Quantum Diffusion CO₂ Capture Optimization NISQ Hardware Validation DAC System Design Gatineau · Ottawa · uOttawa
The Pipeline

From atoms to candidates
in four steps

Our hybrid pipeline encodes molecular structure into quantum states, generates novel candidates, and validates them against multi-objective property targets — all in a single automated workflow.

Molecular Encoding
SMILES strings and molecular graphs are featurized using RDKit and encoded into quantum states via variational quantum circuits (VQCs) on PennyLane.
Quantum Latent Space
A Quantum Variational Autoencoder (Q-VAE) embeds molecules into a quantum-enhanced latent space, improving expressivity for conditional generation tasks.
Generative Sampling
Novel molecular candidates are decoded from the quantum latent space using classical transformer/LSTM decoders with multi-objective reward shaping (QED, SA, logP).
Multi-Objective Screening
Candidates are filtered against synthesizability, ADMET, docking surrogates, and CO₂ adsorption selectivity — shortlisting only the most viable structures for synthesis.
Technology

Four layers.
One unified system.

Layer 01 — Core Research
AI Sorbent Discovery
Graph neural networks trained on DFT/GCMC simulation data predict CO₂ uptake, N₂ selectivity, and regeneration enthalpy across millions of MOF and zeolite candidates — reducing discovery from years to days.
# Screen 10,000 candidate structures
model.screen(candidates, targets={
  "CO2_uptake": > 4.2, # mmol/g
  "selectivity": > 15,
  "regen_dH": < 45 # kJ/mol
})
Layer 02 — Optimization
DAC Process Digital Twin
A real-time ML model continuously optimizes temperature, pressure, flow rates, and regeneration cycles for deployed DAC systems — targeting 20–40% reduction in specific energy consumption.
twin.optimize(system, objective="min_energy")
# → ΔE: -31.4% vs baseline
# → CO₂ purity: 99.2%
# → throughput: +18% tonne/day
Layer 03 — Systems
DAC System Engineering
Co-optimized hardware-software DAC system design integrating sorbent contactors with electrochemical CO₂ separation modules. AI-driven techno-economic modelling from first principles to CAPEX/OPEX targets.
Layer 04 — Novel Capture
Electrochemical CO₂ Capture
Electroswing adsorption and pH-swing electrodialysis cells for electricity-driven CO₂ separation with zero thermal regeneration penalty — purpose-built for Québec's low-cost hydroelectric grid.
Performance

Numbers that matter

100×
Faster than experimental screening
ML-guided screening evaluates 12,000+ candidate structures in hours vs. years of conventional laboratory iteration.
40%
Reduction in DAC energy cost
Digital twin process optimization targets a 20–40% reduction in specific energy consumption (kWh/tonne CO₂) — the dominant driver of DAC economics.
4
Integrated technology layers
From quantum molecular generation through electrochemical capture — a complete full-stack system built to be deployed, not just published.
Built On

Best-in-class toolchain

⚛PennyLane
🔬Qiskit
🧬RDKit
🤖PyTorch
⚗DeepChem
☁Azure ML
🌐IBM Quantum
📊XGBoost
🔗DGL-LifeSci
🧲GCMC Sim
🏗Cirq
🐍Braket SDK
Founders

Built by researchers,
for real problems.

J
Jean-François Cadieux
CEO · AI Engineer
AI engineer specializing in LLM systems, Azure cloud architecture, and production ML pipelines. Leads product strategy, AI engineering, and grant development. Based in Gatineau, Québec.
Azure ML LLM Systems DSPy Québec
A
CTO · CO₂ Researcher
CTO · Materials Scientist
Postdoctoral researcher in CO₂ capture chemistry at the University of Ottawa. Active R&D in solid sorbent materials for direct air capture with deep expertise in adsorption science and electrochemical capture methods.
CO₂ Capture Sorbent Chemistry uOttawa Electrochemistry
Early Access

Ready to accelerate
your discovery?

We're onboarding a small group of early research partners — DAC developers, materials scientists, and industrial CCUS operators. Join the waitlist.

No spam. We'll reach out personally. NRC IRAP · Mitacs · NRCan affiliated.