Alessio Valentini Ph.D.
Software engineer and computational chemist with 20+ publications and 2,400+ citations, combining deep scientific expertise with modern software engineering. Experienced in building molecular modeling platforms and ML pipelines, and in transforming academic prototypes into production-ready, high-performance tools for industry innovation.
My resume
Some of my work!
My background
In my career I have continuously delivered high level software to solve chemistry problems. I am an eager learner and I love to tackle new challenges.
Principal Engineer
- Scaled electronic-structure based featurization pipeline to production-ready CPU/GPU workflow (Docker, Snakemake, Kubernetes on GCP), achieving 200x cost savings and 400x faster runtimes
- Migrated ML featurization workflow to structured database system (SQLAlchemy, Pydantic) ensuring reproducibility and clean data lineage
- Designed custom chemistry tooling (Python, RDKit) to generate and align novel oligomer structures
- Delivered SELEX data-to-dashboard pipeline (Pandas, Streamlit, Dockerized on GCP) adopted by wet-lab groups
- Re-engineered ML training pipelines into modular MLOps framework with metadata/versioning and standardized interpretability
- Developed force-field optimization package, translating ad-hoc cluster workflows into reproducible Kubernetes pipelines
Research Associate
- Led development of Reactions Template Studio, a graph-theory platform for automated exploration of chemical synthesis pathways under ONR MURI grant
- Implemented polarization of QM/MM force fields by interfacing C++ and Fortran codebases, enabling GPU acceleration
- Launched Alexa-powered cloud application for quantum chemical calculations via voice
Postdoctoral Fellow
- Engineered full Python/Cython package for quantum wavepacket propagation with C-like performance
- Analyzed and visualized terabytes of raw data (NumPy, pandas, Jupyter, Matplotlib, Blender) producing insights into nonlinear quantum phenomena
Postdoctoral Fellow
- Streamlined QM/MM workflows by automating complex input generation, reducing setup time from days to hours
- Integrated multiple chemistry packages (ProPKA, GROMACS, Tinker) for high-level objectives like PDB screening and mutant generation
PostDoc
Streamlined the modelling of photoactive proteins to automatize the study of their excited state dynamic behavior (fluorescence and non radiative decay).
PhD in computational chemistry
I got my PhD in Spain, Alcala de Henares. I focused on surface hop semiclassical molecular dynamics applied to natural and synthetic photoactive systems.
Master in Chemistry
I got my master degree in Italy, University of Siena (hosting lab in Bowling Green state university, OH). My research was on the automatic generation of QM/MM models for photoactive proteins.
Bachelor in Chemistry
I got my batchelor degree at the University of Siena (hosting lab University of Wageningen, NL). I studied the prediction of protein folding using bioinformatics tools.
My tools
Programming: Python, C++, Bash, CUDA, Cython, pybind11, Haskell, Fortran Data & Analytics: Pandas, NumPy, Pydantic, BigQuery, PostgreSQL, SQLAlchemy, Alembic, JupyterLab, NetworkX Visualization & UI: Matplotlib, Seaborn, Plotly, D3.js, Streamlit, Blender Machine Learning: PyTorch, TensorFlow, scikit-learn, LightGBM Infrastructure: AWS, GCP, Docker, Kubernetes, GitLab CI/CD, GitHub Actions, Snakemake Computational Chemistry: GROMACS, AMBER, RDKit, OpenEye, Tinker, OpenMolcas, TeraChem, ORCA, xtb