Rafael-michael Karampatsis
AI Decision Science Consultant at Accenture Greece
Akharnaí, Attiki, Greece
Summary
👤
Senior🎓
Top SchoolRafael-michael Karampatsis is an AI Decision Science Consultant with 11 years of experience applying probabilistic machine learning and knowledge representation to real-world problems across finance, consultancy, and academia. He holds a PhD and postdoctoral experience from the University of Edinburgh, where his work spanned explainable AI, program repair, NLP for software engineering, and machine learning for source code. Rafael has moved research into practice through roles at NatWest, DEUS EX MACHINA and now Accenture Greece, translating complex probabilistic models into decision-facing solutions. He combines rigorous academic foundations with hands-on data science in industry, often focusing on interpretable models for social media analysis and software engineering tasks. Notably, his background in program repair and membership in specialized ML for code groups signals an uncommon depth in applying ML to developer tooling and code understanding.
11 years of coding experience
9 years of employment as a software developer
Doctor of Philosophy (Ph.D.), Data Science, Machine Learning for Software Engineering, Program Repair, Natural Language Processing, Doctor of Philosophy (Ph.D.), Data Science, Machine Learning for Software Engineering, Program Repair, Natural Language Processing at The University of EdinburghMaster's Degree, Computer Science, 9.4 (Excellent, with distinction), Master's Degree, Computer Science, 9.4 (Excellent, with distinction) at Athens University of Economics and Business
English, Greek, French
Github Skills (77)
byte-pair-encoding10
mining10
byte10
vocabulary10
arxiv10
encoding9
segmentation9
modeling9
nlp8
hosts8
natural-language-processing8
clustering6
displays6
language-model5
word-segmentation5
Programming languages (5)
Github contributions (5)
mast-group/OpenVocabCodeNLM
Oct 2018 - Feb 2020
Contains the code for our ICSE 2020 paper: Big Code != Big Vocabulary: Open-Vocabulary Language Models for Source Code and for its earlier pre-print: Maybe Deep Neural Networks are the Best Choice for Modeling Source Code (https://arxiv.org/abs/1903.05734). This is the first open vocabulary language model for code that uses the byte pair encoding algorithm (BPE) to learn a segmentation of code tokens into subword units.
Contributions:1 release, 136 commits, 3 PRs in 1 year 4 months
tokenssubword-unitsnatural-language-processingvocabularylanguage-model
mast-group/mineSStuBs
Feb 2019 - Feb 2020
Hosts our tool for mining simple "stupid'' bugs (SStuBs).
Contributions:37 commits, 2 PRs, 37 pushes in 1 year
hostsmining
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.