$whoami

TL;DR you can find my boring CV here.

Hey there, I'm Vladimir, or Vlad in short. I am the author of iSquared. I am a Swiss based Data Scientist.

I have a passion to help my readers understand science through visualisations. That's the best way to digests any concept before delving into the depths.

Following this passion, I created iSquared. It is a space with a goal to enrich your knowledge. If you want to know more about my professional background, check the content below.

To hear more from me and the content I create on iSquared follow me on my social media profiles:

Experience

Full time positions

This list includes all full time positions with an indefinite period of duration.

	Senior Data Scientist Frontiers Mid August 2021 - Present Lausanne, Switzerland
	Working on Recommendation Systems, Natural Language Processing Applications, Information Retrieval and Knowledge Graphs. Main activities: Building Milvus and Vertex AI vector databases of indexed transformer-based text embeddings and serving them for different purposes: recommendations and similarity search. Building, training and evaluating production-ready recommendation systems using PyTorch and Vector Databases for finding the best publications, researchers, reviewers. Training, evaluating, and making production-ready named-entity recognition (NER) systems using spaCy and Huggingface ecosystems. Building production-ready named-entity linking (NEL) systems using high-quality handcrafted bio-medical ontologies and knowledge bases such as SNOMED, MEsH and NCIT. Retrieving data from different sources such as Google Big Query, GraphQL services and making it actionable. Manually handcrafting and labeling NLP data using Prodigy and similar labeling tools. Demoing and integrating the developed solutions in a production-ready environment using Docker, Fast API, and Streamlit. Responsible for organizing tech talks within a team of 50 people in order to break the knowledge silos and reduce the technical debt. Skills: NLP Knowledge Graphs Recommendation Systems Python
	Machine Learning Engineer iGenius, the AI company June 2018 - Mid August 2021 Lausanne, Switzerland
	Work on Crystal: an AI-based virtual advisor for data intelligence. Main activities: Research in deep learning for natural language and speech processing Develop and train state of the art models using TensorFlow and PyTorch for: text-to-speech, named-entity recognition and linking, knowledge graphs, text summarization Demoing and integrating the developed Machine Learning models in a production-ready environment using Docker Composer, Fast API, Postgres SQL, and other Web technologies (Flask, Django) Developing tools in Python for robust objective and subjective evaluation of the developed models as per the established methods in the scientific literature; Creating high-quality handcrafted datasets by means of crowdsourcing using Amazon Mechanical Turk Interviewing prospective candidates for the Machine Learning Engineer role Write high-quality technical documentation Continuous training and education in related topics. Skills: Deep Learning NLP Signal Processing Python TensorFlow PyTorch

Internships

This list includes the full time internships with a predetermined period of duration.

	Master Thesis Project Swisscom September 2017 - March 2018 Lausanne, Switzerland
	Work on my master thesis titled: Building Advanced Dialogue Managers for Goal-Oriented Dialogue Systems Main activities: Research in Goal-Oriented Chatbots based on Reinforcement Learning Create and run experiments to verify or reject our hypothesis Data sets collection and creation Write technical reports and scientific papers Skills: Reinforcement Learning NLP Python LaTeX
	Artificial Intelligence Intern Philip Morris International February 2017 - July 2017 Lausanne, Switzerland
	Main activities: Comparative study and benchmark various Deep Learning Frameworks: TensorFlow, Torch, Theano, MXNet, Café Experiment and test different types of Deep Neural Networks Build and demonstrate Goal-Oriented Chatbot solutions Run A/B tests on the company website for better user engagement Skills: Deep Learning Reinforcement Learning NLP Python LaTeX

Education

Academic Qualifications

	Masters Degree in Computer Sciences École polytechnique fédérale de Lausanne (EPFL) September 2015 - March 2018 Lausanne, Switzerland
	Pursued my master studies with an emphasis on Machine Learning. Language of the studies: English. Relevant courses: Pattern Recognition and Machine Learning (PCML) Unsupervised and Reinforcement Learning in Neural Networks Distributed Information Systems Advanced Database Systems Final Master thesis: Building Advanced Dialogue Managers for Goal-Oriented Dialogue Systems PDF
	Bachelor Degree in Computer Sciences Faculty of Computer Science and Engineering (FCSE) September 2011 - July 2015 Skopje, Macedonia
	Language of the studies: Macedonian with most of literature in English. Relevant courses: Probability and Statistics Intro to Artificial Intelligence Algorithms and Data Structures Web and Mobile Programming Accomplishments Dean’s list for top 10 out of 600 students Scholarship from Ministry of Education for top 10% Computer Science students Final Bachelor Thesis: Transmission of Audio Files using Random Codes Based on Quasigroups Slides

Studentships

This list includes the jobs I had while being a full time student.

	Open Source Student Developer Google Summer of Code May 2017 - August 2017 CERN, Geneva, Switzerland
	Work remotely for the CERN-HSF umbrella organization. Main activities: Design the architecture for the deep learning module in ROOT Develop the foundations of the deep learning module using C++ Develop a Convolutional Neural Network using C++ Skills: Software Engineering Software Design C++ Links: Code Report
	Research Scholar Rigorous System Design Lab (RiSD) September 2015 - January 2017 Lausanne, Switzerland
	Work part time as part of the Research Scholar Program. Main activities: Extend the functionalities of BIP (component-based framework) Design a tool for graphical modeling of BIP architectures using Sirius Develop a complex tool for combining constraint-satisfying components using Java Skills: Software Verification Java SAT Solvers Links: Code
	Summer Student CERN June 2014 - August 2014 Geneva, Switzerland
	Take part of the Summer Student Programme. Main activities: Attend lectures: training in physics, computer engineering and mathematics Develop an Android application as a notification tool Skills: Java Android Math Physics Links: Code Report

Publications

During my studies I succeeded to publish few papers, some on top tier conferences.

Goal-Oriented Chatbot Dialog Management Bootstrapping with Transfer Learning
Vladimir Ilievski, Claudiu Musat, Andreea Hossman, Michael Baeriswyl
In Proceedings of the 27th International Joint Conference on Artificial Intelligence
PDF Bibtex Poster Slides

Submodularity-Inspired Data Selection for Goal-Oriented Chatbot Training Based on Sentence Embeddings
Mladen Dimovski, Claudiu Musat, Vladimir Ilievski, Andreea Hossman, Michael Baeriswyl
In Proceedings of the 27th International Joint Conference on Artificial Intelligence
PDF Bibtex Poster

Gaussian channel transmission of images and audio files using cryptcoding
Verica Bakeva, Aleksandra Popovska-Mitrovikj, Daniela Mechkaroska, Vesna Dimitrova, Boro Jakimovski, Vladimir Ilievski
In IET Communications Journal
PDF Bibtex

Performances of Random Codes Based on Quasigroups for transmission of audio files
Aleksandra Popovska-Mitrovikj, Vladimir Ilievski and Verica Bakeva
IEEE 23rd Telecommunications Forum Telfor (TELFOR)
PDF Bibtex

MOOCs

List of online classes I completed:

	NLP with Classification and Vector Spaces Learned objectives: Classification: Logistic Regression and Naive Bayes Vector Space Models Sentiment Analysis Document Search Certificate
	NLP with Probabilistic Models Learned objectives: Autocorrecting algorithms Part of Speech Tagging Hidden Markov Models and the Viterbi algorithm N-gram language model and autocompletion Word Embeddings Certificate
	NLP with Sequence Models Learned objectives: Sentiment Analysis with Neural Nets RNN Language Models Named Entity Recognition using LSTMs Duplicate Question Detection with Siamese Nets Certificate
	NLP with Attention Models Learned objectives: Neural Machine Translation Text Summarization using Multi-head Attention Question-Answering using BERT Chatbots with Local Sensitive Hashing Attention Certificate
	Statistical Inference Learned objectives: Intro to probability Modelling with different distributions Asymptotics Confidence Intervals and Hypothesis Tests Power and P-values Certificate
	Improving Your Statistical Inferences Learned objectives: Frequentism: correct interpretation of the P-values Bayesian thinking: prior belief and likelihoods Pre-registration and multiple comparisons Effect sizes: significant, but how much? Correct interpretation of the confidence intervals Certificate
	SQL for Data Science Learned objectives: SQL refresher Using SQL for Data Science
	Big Data Analysis: Hive, Spark SQL, DataFrames and GraphFrames Learned objectives: Hive SQL Spark SQL and DataFrame API Spark Optimizations
	Digital Signal Processing Learned objectives: Vector Spaces Fourier Analysis Filters and Filter Design Sampling and Quantization

Languages

English: level C1. Full working proficiency;
French: level B2/B1, DELF B1 certified. Limited working proficiency;
German: level A1/A2. Basic knowledge;
Macedonian: native language.

Accomplishments

Dean's list at the Faculty of Computer Science and Engineering (FCSE) for top 10 out of 600 students
Scholarship from Ministry of Education for top 10% Computer Science students in Macedonia
Bronze medals at the Macedonian Mathematics Olympiad in 2010 for top 20 out of 400 contestants
Second prize at the National Competition in Mathematics in 2010

Hobbies and Activities

Hackathons

I am a huge Hackathon lover, I have participated to many of them. Sometimes just for fun, sometimes competing. Here is a list of selected Hackathons:

2014	CERN Webfest: mobile app for terrain elevation estimation The PORT: Terrain Elevation determination
2016	HackZurich: worked generally on chatbots. LauzHack: worked on a chatbot that predicts train latency. GLAMHack: worked generally on Named-Entity recognition.
2017	HackZurich: worked on object detection using Raspberry Pi. NYUADHack: app to connect refugees and the local community. Winners. OpenFood Hack: worked on app for OCR-ing receipts. One of the winners. StartHack: detecting travelling trends through Twitter. Finalists.
2018	LauzHack: worked on a fraud detection system. Challenge winners.
2019	Arkathon: worked on an app for pose detection for better exercising. BlueArk Hack: better water waste in the households.

Hobbies

Running: ran 2750 kilometers in 2023, half-marathon personal best 1:33:28. Aiming to run under 1:30:00

Aviation: beside the computer sciences I am passionate about the commercial airplane models along with their jet engines, their design, engineering as well as the business models of the airliners.

Hiking: hiking sometimes through the beautiful Swiss Alps.

Vladimir Ilievski

Subscribe