resume.tex

\documentclass[journal]{IEEEtran}
\input{include/preamble}

\title{Résumé of Ivan Ukhov}
\author{
  Ivan Ukhov\\
  \link[,]{mailto:ivan.ukhov@gmail.com}{ivan.ukhov@gmail.com}
  \link[,]{https://blog.ivanukhov.com}{blog.ivanukhov.com}
  \link[,]{https://research.ivanukhov.com}{research.ivanukhov.com}\\
  GitHub: \link[,]{https://github.com/IvanUkhov}{IvanUkhov}
  LinkedIn: \link[,]{https://www.linkedin.com/in/IvanUkhov/}{IvanUkhov}
  X: \link{https://x.com/IvanUkhov}{IvanUkhov}
}

\begin{document}

\maketitle

\begin{abstract}
In this paper, we summarize Ivan Ukhov's education and career as well as other
aspects related to his professional and personal growth. The objective is to
make it easier for potential employers to obtain a sufficiently comprehensive
understanding of the candidate's profile and subsequently decide if and how to
proceed with the recruitment process. However, the effectiveness of this
presentation format remains an open research question.
\end{abstract}

\begin{IEEEkeywords}
  Airflow,
  Cloudflare,
  computer science,
  data science,
  Git,
  GitHub,
  Google Cloud Platform,
  GraphQL,
  JavaScript,
  Kubeflow,
  Kubernetes,
  machine learning,
  OAuth~2.0,
  open source,
  probability theory,
  Python,
  research,
  Rust,
  software engineering,
  SQL,
  statistics,
  teaching,
  TensorFlow,
  Terraform.
\end{IEEEkeywords}

\bstctlcite{IEEEexample:BSTcontrol}

\section{Introduction}

\lettrine[findent=0.4em, nindent=0em]{\textbf{T}}{alent acquisition} becomes
increasingly more difficult. There are many candidates with their unique
backgrounds, abilities, and preferences, making it challenging to navigate.
Under these circumstances, it is important to be able to reasonably fast make a
sufficiently accurate evaluation if the candidate in question would be a good
fit for the job at hand. Ivan Ukhov is one such candidate, and this paper
presents his profile to assist in the aforementioned evaluation process. In what
follows, we shall refer to Ivan as the candidate.\footfloat{\linkmarker An omega
superscript indicates the presence of a link, which is sadly not accessible in
print form due to the limitations of the medium.}

The subsequent sections go over several topics in the order of decreasing
importance. \sref{interests} and \ref{section:skills} describe the candidate's
professional interests and technical skills, respectively, which are arguably
the ones dictating the drive and the ability to follow through. \sref{work},
\ref{section:teaching}, and \ref{section:learning} present his experience in
terms of working, teaching, and learning, respectively. In \sref{education}, the
candidate's formal education is summarized. \sref{publications} and
\ref{section:projects} list some of his scientific publications and open-source
projects, respectively. In \sref{personal}, a glimpse into the candidate's
personal life is given. Lastly, \sref{conclusion} concludes the paper.

\section{Professional Interests} \slab{interests}

We begin with the candidate's professional interests. They revolve around making
business decisions under uncertainty and solving the accompanying engineering
problems. This includes choosing directions based on evidence combined with
prior knowledge, answering questions via inferential and predictive modeling,
optimization of processes by means of learning from data, and development of
data products. The underlying disciplines of interest are data science, machine
learning, software engineering, and statistics.

\section{Technical Skills} \slab{skills}

Apart from the general, strong knowledge of software engineering, statistics,
and machine learning, there are several specific methodologies and technologies
the candidate has come across and acquired throughout his education and career:
strong knowledge of Rust and Python, including their ecosystems; good knowledge
of C++, Go, R, and shell scripting; strong knowledge of SQL; familiarity with
JavaScript, HTML, and CSS; strong knowledge of TensorFlow; strong knowledge of
Google Cloud Platform, including Kubernetes Engine and Vertex AI; strong
knowledge of Docker and Terraform; good knowledge of GraphQL; strong knowledge
of Airflow and Kubeflow; strong knowledge of GitHub, including GitHub Actions;
strong knowledge of software testing and continuous integration and deployment;
proficiency in version control with Git; and proficiency in Vim and LaTeX.

\section{Work Experience} \slab{work}

There have been three non-academic employers: LeoVegas (2018), Voi (2022), and
The Type Founders (2023). The corresponding roles are given in reverse
chronological order, which is also the one used in the remainder of the paper.

\date{February 2023--Present} \emph{Staff Engineer at The Type Founders}: \sep
Established a platform for managing and serving a large library of typefaces
(Rust, Tokio, GraphQL, Google Cloud Platform, GitHub Actions). \sep Built a
system for identifying visually similar typefaces (convolutional autoencoders).
\sep Explored generation of typefaces (transformers). \sep Analyzed the state of
the web with respect to the usage of fonts, as part of the Web Almanac by the
HTTP Archive (BigQuery).

\date{February 2022--January 2023} \emph{Staff Machine Learning Engineer at
Voi}: \sep Established a platform for developing and running machine-learning
projects, used throughout the company (Google Cloud Platform, GitHub Actions).
\sep Participated in other projects led by the team, including inferential
modeling based on sensor data from IoT units and image classification
(TensorFlow) with object detection (PyTorch) in constrained environments (mobile
devices, Raspberry Pi). \sep Participated in the recruitment of machine-learning
engineers. \sep Mentored colleagues in machine learning and software
engineering.

\date{November 2019--February 2022} \emph{Head of Data Science at LeoVegas}:
\sep Led the team, ensuring clarity and relevance of direction. \sep Held group
and individual meetings. \sep Performed planning and setting of common and
individual objectives.

\date{February 2018--February 2022} \emph{Data Scientist at LeoVegas}: \sep
Developed an experimentation platform for democratization and decentralization
of hypothesis testing (Google Cloud Platform, Bayesian statistics). \sep
Developed a predictive model for identifying individuals who are at risk of
gambling addiction (gradient boosting, recurrent neural networks). \sep
Developed an inferential model and a natural-language-processing model for
analyzing the results of customer surveys (Bayesian statistics, recurrent neural
networks). \sep Developed a data-driven artificial environment for optimizing
promotion campaigns via reinforcement learning (feedforward neural networks).
\sep Developed a platform for data-science projects, including pipelines for
data preparation, frameworks for predictive modeling, and infrastructures for
scheduling and serving predictive models (Google Cloud Platform). \sep
Participated in other projects led by the team, including causal inference in
observational studies, churn prediction, high-value prediction, inference of
spillover in marketing, lifetime-value prediction, marketing mix modeling, and
recommender systems. \sep Participated in the recruitment of data scientists.
\sep Mentored team members in data science, machine learning, and software
engineering; see also \sref{teaching}. \sep Led research projects on the topic
of problem gambling in collaboration with the Karolinska Institute, Stockholm
University, and Nottingham Trent University.

\section{Teaching Experience} \slab{teaching}

In this section, we summarize the candidate's experience in terms of helping
others to grow professionally.

\date{2022} Supervised two Master's students at Voi on the topic of identifying
irresponsible, unsafe usage of electrical vehicles via sensor readings by means
of outlier detection.

\date{2018--2019} Supervised two Master's students at LeoVegas on the topic of
identifying problem gamblers using recurrent neural networks and on the topic of
managing promotion campaigns using reinforcement learning.

\date{2011--2017} Supervised one Ph.D. student and five Master's students at
Linköping University in machine learning, software engineering, and computer
systems.

\date{2011--2017} Assisted in the following undergraduate courses at Linköping
University: \link[,]{https://studieinfo.liu.se/kurs/tddb84/ht-2018}{Design
Patterns} \link[,]{https://www.ida.liu.se/~TDDD25/}{Distributed Systems}
\link[,]{https://www.ida.liu.se/~TDDI08/}{Embedded Systems Design}
\link[,]{https://www.ida.liu.se/~TDDD04/}{Software Testing} and
\link[.]{https://www.ida.liu.se/~TDTS07/}{System Design and Methodologies}

\section{Learning Experience} \slab{learning}

In this section, we go over several noteworthy courses that the candidate has
passed throughout the years.

\date{2022} Passed the Machine Learning specialization on Coursera (an updated
classic by Andrew Ng) and hosted a ``book'' club for aspiring machine-learning
practitioners at Voi.

\date{2018--2019} Passed the following specializations on Coursera: Deep
Learning, Machine Learning with TensorFlow on Google Cloud Platform, and Data
Engineering.

\date{2011--2017} Passed the following graduate courses at Linköping University:
Advanced Data Models and Databases, Bayesian Learning, Data Mining and
Statistical Learning, Distributed Systems, Gaussian Random Processes, Malliavin
Calculus and Stochastic Integration, Natural Language Processing, Neural
Networks with Applications to Vision and Language, Probability Theory, Real-Time
and Embedded Systems, Stochastic Optimization, and Stochastic Processes.

\section{Academic Qualifications} \slab{education}

In this section, the candidate's formal equation is given, which can be broken
down into two periods: Saint Petersburg, Russian, until 2011 and Linköping,
Sweden, from 2011.

\date{2017} \emph{Doctor of Philosophy in Computer Science}, Embedded Systems
Laboratory, Department of Computer and Information Science, Linköping
University.

\date{2010} \emph{Master of Science in Computer Science} with \textsc{honors},
Department of Information and Control Systems, Peter the Great Saint Petersburg
Polytechnic University.

\date{2010} \emph{Specialist in Business Management} with \textsc{honors},
International Graduate School of Management, Peter the Great Saint Petersburg
Polytechnic University.

\date{2008} \emph{Bachelor of Science in Computer Science} with \textsc{honors},
Department of Information and Control Systems, Peter the Great Saint Petersburg
Polytechnic University.

\section{Selected Scientific Publications} \slab{publications}

In this section, we list several notable scientific publications authored by the
candidate, which were produced during his Ph.D. education; see \sref{education}.

\date{2017} \emph{System-Level Analysis and Design under
Uncertainty}~\cite{ukhov2017d}: In a Ph.D. dissertation, summarized and
unified the individual pieces of research listed below.

\date{2017} \emph{Fine-Grained Long-Range Prediction of Resource Usage in
Computer Clusters}~\cite{ukhov2017b}: Studied resource usage in a
\link[.]{https://github.com/google/cluster-data}{computer cluster} Constructed
an efficient pipeline for data processing and devised a recurrent neural network
for forecasting resource usage. Made use of TensorFlow.

\date{2017} \emph{Fast Synthesis of Power and Temperature Profiles for the
Development of Data-Driven Resource Managers}~\cite{ukhov2017c}: Studied network
traffic in a \link[.]{https://github.com/google/cluster-data}{computer cluster}
Developed an infrastructure for simulating systems processing user requests with
the goal of providing large amounts of synthetic yet realistic data to
facilitate the development of resource managers powered by machine learning.

\date{2017} \emph{Probabilistic Analysis of Electronic Systems via Adaptive
Hierarchical Interpolation}~\cite{ukhov2017a}: Developed an efficient framework
for probabilistic analysis of electronic systems based on adaptive hierarchical
interpolation on sparse grids. Leveraged advanced topics in numerical analysis.

\date{2015} \emph{Temperature-Centric Reliability Analysis and Optimization of
Electronic Systems under Process Variation}~\cite{ukhov2015}: Developed an
efficient probabilistic framework for reliability analysis of electronic systems
using polynomial regression and applied this framework in the context of energy
optimization.

\date{2014} \emph{Probabilistic Analysis of Power and Temperature under Process
Variation for Electronic-System Design}~\cite{ukhov2014b}: Developed an
efficient probabilistic framework for temperature analysis of electronic systems
under process variation based on polynomial regression. Made use of advanced
topics in probability theory and numerical analysis.

\date{2014} \emph{Statistical Analysis of Process Variation Based on Indirect
Measurements for Electronic-System Design}~\cite{ukhov2014a}: Developed an
efficient statistical framework for characterizing variations in parameters of a
technological process based on indirect measurements. Made use of Bayesian
inference.

\date{2012} \emph{Steady-State Dynamic Temperature Analysis and Reliability
Optimization for Embedded Multiprocessor Systems}~\cite{ukhov2012}: Developed a
fast and accurate technique for temperature analysis of multiprocessor systems
under periodic workload and applied this technique in the context of reliability
optimization. Made use of advanced linear algebra.

\section{Selected Open-Source Projects} \slab{projects}

In this section, several of the candidate's open-source projects are given. Each
item below corresponds to an organization on GitHub where the individual
packages can be found.

\emph{\link[:]{https://github.com/stainless-steel}{Stainless Steel
(stainless-steel on GitHub)}} Developed a collection of Rust packages of general
interest in such interrelated areas as linear algebra, probability theory,
statistics, signal processing, and relational databases.

\emph{\link[:]{https://github.com/bodoni}{Bodoni (bodoni on GitHub)}} Developed
a collection of packages in Rust for working with vector graphics and font
formats, including OpenType and Web Open File Format.

\emph{\link[:]{https://github.com/ready-steady}{Ready Steady (ready-steady on
GitHub)}} Developed a collection of Go packages of general interest in such
interrelated areas as linear algebra, numerical integration, interpolation,
probability theory, and statistics.

\emph{\link[:]{https://github.com/markov-chain}{Markov Chain (markov-chain on
GitHub)}} Developed a collection of Rust packages for high-level simulation of
multiprocessor systems with an emphasis on their thermal dynamics, which was
used in research; see \sref{publications}.

\emph{\link[:]{https://github.com/turing-complete}{Turing Complete
(turing-complete on GitHub)}} Developed a collection of Go packages for
high-level simulation of multiprocessor systems with an emphasis on their
thermal dynamics, which was used in research; see \sref{publications}.

\emph{\link[:]{https://github.com/learning-on-chip}{Learning on Chip
(learning-on-chip on GitHub)}} Developed a collection of tools in Rust and
Python for processing, simulation, and prediction of dynamics in a
\link{https://github.com/google/cluster-data}{computer cluster} for research
purposes; see \sref{publications}.

\emph{\link[:]{https://github.com/chain-rule}{Chain Rule (chain-rule on
GitHub)}} Developed a collection of packages used in the
\link{https://blog.ivanukhov.com/}{personal blog} about data science.

\section{Personal Information} \slab{personal}

The candidate considers himself to be reliable, responsible, meticulous, and
organized. He likes to think that he has high standards for the written word and
the written code. He claims to work well both in a team and individually.

The candidate is located in Stockholm, Sweden. He goes to the gym, likes
\link[,]{https://photography.ivanukhov.com/}{photography}
\link[,]{https://blog.ivanukhov.com/}{blogs about software engineering} and
plays the piano, or rather aspires to. He enjoys typography and type design.

\section{Conclusion} \slab{conclusion}

In this paper, we summarized Ivan Ukhov's profile so as to assist potential
employers in evaluating the candidate's relevance for their endeavors. It is
well understood, however, that the actual causal impact of the present format on
the efficiency and effectiveness of the recruitment process remains unknown,
which we leave for future work.

\begingroup
  \bibliographystyle{IEEEtran}
  \bibliography{IEEEabrv,include/bibliography}
\endgroup

\end{document}