Skip to content
Portrait of Tarik Ammouli
Tarik Ammouli, Ph.D.

Executive profile · Paris, France

Tarik Ammouli, Ph.D.

Computational Physical Chemist · Research Software Engineer

I turn the equations of statistical thermodynamics and quantum chemistry into software that runs at scale. From the adsorption of pollutants in zeolites to the electronic structure of thousands of drug-like molecules, my work sits where in-silico discovery meets distributed Cloud/HPC engineering and the discipline of shipping research as a product.

Download CV (PDF)tarik.ammouli@gmail.comRead the monograph
Languages
French · English · Arabic (fluent)
ResearchGate
Tarik-Ammouli

Career odyssey

Four chapters, one through-line: making molecular science run in production.

Each chapter reads as a short case study — the context, the scientific or technical challenge, how it was built, and what it changed.

  1. I

    Paris, France

    Industrialising Drug Discovery at Scale

    Research Software Engineer, Computational Chemistry · Qubit Pharmaceuticals

    Cross-functional bridge between medicinal chemists and software teams. Scaling structure- and ligand-based virtual screening to millions of compounds, building the ab-initio data pipeline behind the company's foundation model, and running resilient Cloud/HPC architectures.

    • Python
    • AWS
    • Kubernetes
    • Docker / Singularity
    • CI/CD
    • RDKit
    • PySCF
    • AutoDock Vina
    • Milvus

    01 · Context

    Qubit Pharmaceuticals, a DeepTech spin-off of CNRS and Sorbonne Université, builds the Atlas platform: physics-grade simulation and AI for drug discovery. Research prototypes had to become reliable, cost-controlled production workflows used daily by medicinal and computational chemists.

    02 · Scientific & Technical Challenge

    Three bottlenecks at once: generating millions of high-fidelity quantum-chemistry data points for machine and deep learning, screening chemical libraries of millions of compounds in reasonable wall-clock time, and keeping heterogeneous cloud and HPC infrastructure resilient and affordable.

    03 · Architecture & Execution

    Designed and delivered end-to-end workflows in close collaboration with chemists, then industrialised them.

    • Built a scalable ab-initio electronic-structure workflow characterising thousands of compounds and producing millions of training data points for the company's foundation model, the backbone of its AI strategy.
    • Designed molecular enumeration and 3D-conformer modelling pipelines with medicinal chemists, improving small-molecule preparation strategies.
    • Scaled structure- and ligand-based virtual screening by integrating parallelised docking engines (speed and binding-score accuracy) and a Milvus vector-search database for similarity searches across millions of compounds.
    • Developed bots that automate data recovery across multiple HPC clusters, cutting operational overhead.
    • Industrialised legacy codebases: refactoring experimental scripts into maintainable services with CI/CD, automated tests, containerisation (Docker, Singularity/Apptainer), code reviews and documentation.
    • Designed and deployed a cloud cost-control strategy (budget tracking, cost/performance arbitration) on AWS.

    04 · Tangible Impact

    Faster drug-discovery cycles and a leaner platform.

    • Infrastructure spend reduced by 20% through the cloud cost-governance strategy.
    • Virtual screening now runs on millions of compounds through distributed compute and automated CI/CD flows.
    • Translated R&D needs into technical roadmaps; mentored interns and onboarded newcomers; provided technical support to internal users of the Atlas platform.
    Deep divesAb-initio data pipeline for a chemical foundation modelSmall-molecule preparation workflows for chemical librariesVirtual screening across millions of compoundsCloud cost governance for a compute-heavy platform

  2. II

    Marseille, France

    Quantum Interfaces & Nanomaterials

    Postdoctoral Researcher · CINaM – CNRS UMR 7325, Aix-Marseille Université

    Boosting the chiral selectivity of single-walled carbon nanotubes during catalytic growth: DFT, Bader charge analysis and tight-binding models of nanotube–nickel interfaces to guide CCVD synthesis.

    • DFT
    • VASP
    • Bader analysis
    • Tight-Binding
    • SWCNT
    • Catalysis
    • HPC

  3. III

    Mulhouse, France

    Multi-Scale Simulation & Environmental Remediation

    Ph.D. Researcher in Physical Chemistry · IS2M – Université de Haute-Alsace (CNRS UMR 7361)

    Coupling periodic DFT, molecular mechanics and configurational-bias Monte Carlo with synthesis and characterisation (XRF, XRD, TGA/DSC, manometry) to design cationic faujasite zeolites that capture persistent organic pollutants in the presence of water.

    • Periodic DFT
    • VASP
    • GULP
    • Towhee (CBMC / GEMC)
    • XRD · XRF
    • TGA / DSC
    • Manometry
    • Teaching

  4. IV

    Casablanca, Morocco

    Academic Foundations & Distinctions

    M.Sc. & B.Sc. in Physics · Université Hassan II de Casablanca

    Master's degree in Physics & New Technologies, graduated first of the class (Major de promotion), after a Bachelor's degree in Physics & Applications.

    • Statistical physics
    • Quantum mechanics
    • Numerical methods
    • Materials

Deep dives

Projects and thesis chapters, opened like a reader.

Each card opens a side reader with the scientific problem, the computational and experimental stack, figure slots and the key takeaways.

Publications, conferences & honors

A bibliography built on simulation and experiment.

Peer-reviewed articles, the doctoral thesis and conference contributions. Every entry carries a one-click BibTeX export and a link to the DOI or open-access record.

  1. 2026Journal article

    Adsorption of polychlorinated aromatics in potassium exchanged faujasite: A molecular simulation study

    Tarik Ammouli, Jean-Louis Paillaud, Habiba Nouali, Marie-Christine Hanf, Régis Stephan, Philippe Sonnet, Irena Déroche

    Microporous and Mesoporous Materials 413, 114272

    A combination of DFT and Monte Carlo simulation explores the adsorption of dichlorobenzene isomers in potassium-exchanged X-type faujasite: preferential locations, hydrogen bonds and K⁺ electrostatics, absolute isotherms and adsorption heats versus loading, and the redistribution of K⁺ cations towards the supercage.

    DOI 10.1016/j.micromeso.2026.114272HAL hal-05848983DFTMonte Carlo molecular simulationPotassium exchanged X-type faujasiteAdsorption
    DOI
  2. 2021Journal article

    Insights into Water Adsorption in Potassium-Exchanged X-type Faujasite Zeolite: Molecular Simulation and Experiment

    Tarik Ammouli, Jean-Louis Paillaud, Habiba Nouali, Régis Stephan, Marie-Christine Hanf, Philippe Sonnet, Irena Deroche

    The Journal of Physical Chemistry C 125(35), 19405–19416

    Grand canonical and canonical Monte Carlo simulations combined with DFT and experiment describe water adsorption in KX faujasite from low loading to saturation, the preferential coordination of K⁺ at sites III/III′ over site II in the supercages, and the hydration-induced migration of cations.

    DOI 10.1021/acs.jpcc.1c03593Water adsorptionFaujasiteGrand canonical Monte CarloDFT
    DOI
  3. 2022Ph.D. thesis

    Adsorption de dichlorobenzènes par des zéolithes cationiques en présence d'eau

    Tarik Ammouli

    Ph.D. thesis, Université de Haute-Alsace – Mulhouse (ED 182)

    NNT 2022MULH4688 · defended 3 February 2022 · directors Philippe Sonnet & Régis Stephan

    HAL tel-03888267DichlorobenzeneDFTMonte CarloHydration rate
    Open access
  4. 2023Poster

    A DFT study of the nanotube-catalyst charge transfer during the CVD growth

    Tarik Ammouli, Christophe Bichara

    NT23 – 23rd International Conference on the Science and Applications of Nanotubes and Low-Dimensional Materials · Arcachon, France · 4 – 9 June 2023

    Open access
  5. 2021Poster

    Removal of Persistent Organic Pollutants (POPs) by cationic zeolites under humid conditions: a coupled in-silico / experimental investigation

    Tarik Ammouli, Jean-Louis Paillaud, Philippe Sonnet, Régis Stephan, Marie-Christine Hanf, Irena Deroche

    FEZA 2021 – 8th Conference of the Federation of European Zeolite Associations · Virtual (organised from Brighton, UK) · 5 – 9 July 2021

  6. 2022Oral presentation

    Influence of extra-framework cations and hydration rate on adsorption properties of polychlorinated aromatics in zeolite

    Tarik Ammouli, Jean-Louis Paillaud, Philippe Sonnet, Régis Stephan, Marie-Christine Hanf, Irena Deroche

    Matériaux 2022 – Conférence internationale de la Fédération Française des Matériaux · Lille, France · 24 – 28 October 2022

Honors and certifications

  • Best Poster Award

    Young Scientist Day, IS2M – Institut de Science des Matériaux de Mulhouse

    Dec 2019

  • CECAM Molsim20 Fellow

    Summer school on molecular simulation, Van 't Hoff Institute for Molecular Sciences, University of Amsterdam

    Jan 2020

  • Major de promotion

    M.Sc. Physics & New Technologies, Université Hassan II de Casablanca

    2018

CertificationsPharma: BI for Biopharma · Pharma Industry Fundamentals · Cheminformatics & MedChemData & Infrastructure: AWS Cloud · Kubernetes · Apache Airflow 3 · VirtualizationSoftware Architecture: Data Architectures · Kibana · UML System Design

Dual-core expertise

Two disciplines, held together.

Strategy without a working system is a slide deck; a system without governance is a cost centre. The practice is to run both.

α

DeepTech Strategy & Leadership

Turning research into products that ship, with the budget to match.

R&D to roadmap
Translating scientific needs into technical roadmapsProduct industrialisationScientific storytelling for non-specialists
Cross-functional alignment
Bridge between medicinal chemists and engineeringStakeholder alignmentMentoring interns & onboarding
Delivery & governance
Agile delivery (Asana, Jira, Notion)Cloud cost governance (20 % lower infrastructure spend)Managing uncertainty in R&D projects
Architecture & design
UML & ERD system designData architecturesObservability-driven operations

β

Computational Chemistry & Cloud/HPC Engineering

From the Schrödinger equation to a container running on a cluster.

Languages & frameworks
PythonGoRubyFortranBashSQLFastAPIDjangoRailsStreamlit
Cloud & data
AWS (EC2, EKS, S3, RDS, Lambda, ParallelCluster)PostgreSQLElasticsearchMilvusRedis
DevOps & CI/CD
GitDockerKubernetesSingularity / ApptainerGitLab CIGitHub ActionsArgoCD
Observability & data engineering
PrometheusGrafanaELK StackAirflow
Cheminformatics & virtual screening
RDKitChemAxonOpenBabelAutoDockVina / VinaXB
Atomistic modelling
VASPPySCFGULPTowheeTinker-HP
Analytics & visualisation
Materials StudioASEVASPKITVMDPyMOLVESTADataWarrior

Contact & collaboration

Let's build what comes after the prototype.

Open to R&D partnerships, DeepTech advisory and conversations with founders and operators who need science that runs in production.

Direct coordinates

tarik.ammouli@gmail.com

Paris, France