Skip to content
View skafai's full-sized avatar
🎯
Focusing
🎯
Focusing

Highlights

  • Pro

Organizations

@chronocratic @relflow

Block or report skafai

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
skafai/README.md

Wadie Skaf: AI Research Scientist, Developer, and Passionate Technologist

I am an AI Research Scientist and PhD candidate in Computer Science at RWTH Aachen University, researching automated machine learning (AutoML), uncertainty quantification, and self-supervised representation learning for time series. My work focuses on developing explainable, trustworthy, and uncertainty-aware models that are less dependent on labelled data and manual algorithm design.

I am interested in building open-source packages that improve AI research, as well as Rust-based tools that make AI more efficient. I also actively contribute to open-source projects.


Research Focus

Area Topics
Time Series Self-supervised representation learning, classification, and forecasting
AutoML Meta-learning, neural architecture search (NAS), and hyperparameter optimization (HPO)
Uncertainty Quantification Conformal prediction

Technical Skills

Core
Machine Learning Deep Learning Statistics Mathematics Scientific Computing Software Engineering Cloud Engineering

Languages
Python C++ Rust JavaScript TypeScript SQL Bash

ML & DL
PyTorch Lightning TensorFlow Keras scikit-learn

Data
Pandas NumPy Matplotlib Seaborn Apache Spark

MLOps & infra
MLflow Docker Linux Slurm AWS Git Apache Airflow

Web
React Next.js Django Tailwind CSS


Projects

Chronocratic
Open-source infrastructure, datasets, and models for time series research
GitHub

→ tslab
Batteries-included lab for time series research
GitHub

→ datasets
Ready-to-use time series datasets and DataModules for PyTorch Lightning
GitHub

→ models
Ready-to-use time series models in PyTorch and Lightning
GitHub


rewrite.rs
Guide and AI skills for rewriting software in Rust
Website GitHub


RATS
High-performance time series augmentation library in Rust, with a Python wrapper (RATSpy)
GitHub


PCGA
Genetic algorithm for controlling label proportions across containers
GitHub


Open Source Contributions

relflow
Open-source model factory that builds and adapts foundation models for typed, nested structured data
GitHub


Main Publications

Skaf, W., Baratchi, M., & Hoos, H. (2025). Time Series Representations Classroom (TSRC): A teacher-student-based framework for interpretability-enhanced unsupervised time series representation learning. Machine Learning.
Paper Code

Moeini, E., Vox, C., Anastacio, M., Skaf, W., Baratchi, M., & Hoos, H. H. (2026). Neural architecture and hyperparameter selection through meta-learning on time series. In Proceedings of the AAAI Conference on Artificial Intelligence (AAAI).
Paper

Skaf, W., & Horváth, T. (2022). Denoising architecture for unsupervised anomaly detection in time-series. In Proceedings of the European Conference on Advances in Databases and Information Systems (ADBIS).
Paper Code


Education

Dr. rer. nat. (PhD), Computer Science (AI) ongoing
RWTH Aachen University
AutoML, uncertainty quantification, time series, and representation learning

M.Sc., Computer Science (AI)
Eötvös Loránd University
Machine learning, deep learning, anomaly detection, and audio signals

B.Sc., Informatics (Software Engineering)
Latakia University
Software engineering, information systems, databases, and mobile apps


Open to Work

I am available for freelance and consulting projects across my skill set, including but not limited to:

  • Machine learning research and development
  • Time series analysis and forecasting
  • Making AI efficient with Rust (fast inference, data pipelines, and tooling)
  • MLOps and AI systems engineering
  • Software and web development

Reach out on LinkedIn or by email.


Links

Website Google Scholar ORCID LinkedIn Email

Pinned Loading

  1. chronocratic/tslab chronocratic/tslab Public

    A batteries-included lab for time series research.

    Python 4

  2. chronocratic/chronocratic-models chronocratic/chronocratic-models Public

    Ready-to-use time series models implemented in PyTorch and Lightning.

    Python 4 1

  3. rewrite-rs/skills rewrite-rs/skills Public

    The most comprehensive agent skills for writing real, idiomatic Rust and porting code from other languages.

    JavaScript 21 3

  4. HyperVectors/RATS HyperVectors/RATS Public

    Rapid Augmentation Library for Time Series data

    Rust 10