Dénes Bartha, Developer in Singapore, Singapore
Dénes is available for hire
Hire Dénes

Dénes Bartha

Verified Expert  in Engineering

Bio

As a Ph.D. student of Computer Science, Dénes has worked as a researcher in Bioinformatics at The University of Tokyo and the National University of Singapore. He has also contributed as a Software Engineer at the Canadian Aviation Engineering (CAE). He greatly enjoys using machine learning techniques in order to solve real-world problems and to help businesses.

Portfolio

Dkatalis
TensorFlow, Google Cloud Platform (GCP), BigQuery, Python, AutoML...
Doctor Anywhere
BigQuery, Python, MySQL, Tableau, Google Cloud Platform (GCP)...
National University of Singapore
Python, C++, Pandas

Experience

  • C++ - 10 years
  • Python - 8 years
  • Artificial Intelligence (AI) - 6 years
  • Machine Learning - 6 years
  • Scikit-learn - 3 years
  • BigQuery - 3 years
  • TensorFlow - 3 years
  • Hardware Development - 3 years

Availability

Full-time

Preferred Environment

Jupyter Notebook, PyCharm, Git, Ubuntu, Google Cloud Platform (GCP), Sensor Data

The most amazing...

...tool that I have made is a DNA data compression/assembler program called Colorgram. It is a Succinct Colored de Bruijn Graph variant.

Work Experience

Senior Data Scientist

2020 - PRESENT
Dkatalis
  • Implemented machine learning pipeline for transaction classification in GCP, orchestrated via Dataflow and Kubeflow. Conducted model training and hyperparameter tuning by Vertex AI AutoML and created custom models using Katib with LightGBM, Bert, and TensorFlow.
  • Developed a machine learning pipeline to detect recurring transactions based on customer and transaction information. The model filled the feature "Plan Ahead" in the app's front end based on the detected recurring transactions.
  • Integrated an in-house built TensorFlow lite YOLO model for automatically detecting IC cards. Added image quality checks to the front end and employed Google Vision Text Recognition for parsing content, applying NLP techniques for result cleansing.
  • Participated in creating an API service that shows various insights to the users in the mobile app via Braze, orchestrated with Kafka, using Redis and PostgreSQL.
  • Established data quality frameworks for our database in BigQuery implemented in Python using Great Expectations and orchestrated via Apache Airflow. Implemented a custom SQL unit testing framework.
Technologies: TensorFlow, Google Cloud Platform (GCP), BigQuery, Python, AutoML, Cloud Dataflow, Kubeflow, Kubernetes, BERT, Natural Language Processing (NLP), Language Models, Apache Airflow, Data Build Tool (dbt), Terraform, Scikit-learn, You Only Look Once (YOLO), Dart, Machine Learning, Artificial Intelligence (AI), PyTorch, Pandas, Google Colaboratory (Colab), Convolutional Neural Networks (CNNs), Computer Vision, Deep Learning, Recurrent Neural Networks (RNNs), Large Language Models (LLMs), Machine Learning Operations (MLOps)

Senior Data Scientist

2019 - 2020
Doctor Anywhere
  • Helped finance, operations, marketing, BD, and doctors for creating/automating reports in Python and MySQL, sending out daily mails automatically from AWS and GCP Linux virtual machines.
  • Automated the integration of a 3rd-party healthcare platform used by our clinics via their API in Python and JavaScript. The platform was missing some CMS functionalities, e.g., setting low stock alerts and calculating cost prices automatically.
  • Implemented a pipeline in Python for pulling data from various sources, including multiple MySQL servers, MongoDB, Microsoft SQL Server, and Firebase into BigQuery.
  • Created long, flat tables and views in BigQuery using Standard SQL and integrated these with Tableau so that other teams could easily access and analyze the data independently.
  • Created an ensemble Random Forest classifier in Python using scikit-learn libraries to predict patient diagnoses based on symptoms, reducing doctors' logging time and filtering unsuitable cases.
  • Estimated patients' claim prices using XGBoost in Python.
  • Optimized medication delivery routes by analyzing rider data and geolocation data using Standard SQL and Python.
Technologies: BigQuery, Python, MySQL, Tableau, Google Cloud Platform (GCP), Amazon Web Services (AWS), JavaScript, Scikit-learn, XGBoost, Machine Learning, Artificial Intelligence (AI), TensorFlow, PyTorch, Pandas, Google Colaboratory (Colab), Computer Vision, Deep Learning, Large Language Models (LLMs), Machine Learning Operations (MLOps), Microsoft Power BI

Researcher

2018 - 2019
National University of Singapore
  • Worked in the bioinformatics laboratory of the Computer Science Department.
  • Created design and implementation of concrete bioinformatical algorithms.
  • Analyzed data and statistics of human and virus DNA.
  • Worked on DNA compression and assembly-related problems.
  • Created Colorgram—succinct colored de Bruijn graph.
Technologies: Python, C++, Pandas

Researcher

2016 - 2017
University of Tokyo
  • Worked in a bioinformatics laboratory. Created theoretical algorithms related to bioinformatical problems.
  • Analyzed mass spectrometry data and implemented and tested various DNA reconstruction algorithms.
  • Created and presented statistics and published results in Acta Cybernetica scientific journal.
Technologies: Python, C++, Pandas

Software Engineer

2014 - 2017
Canadian Aviation Electronics (CAE)
  • Supported the development of the pilot training system by working on both the UI and the back end.
  • Maintained the components by analyzing the customers' data and feedback.
  • Designed and developed a specific communication system for military aircraft.
  • Collaborated (daily) between the Hungarian and Canadian sites.
Technologies: C#, Python, C++, Machine Learning Operations (MLOps)

Data Scientist

2014 - 2014
Nextent Informatics Co.
  • Supported the data collection from the customers.
  • Analyzed data using machine learning techniques.
  • Created statistics.
  • Supported creating the design of mobile application.
  • Participated in the developed mobile application for android.
Technologies: Android, Python, R, Machine Learning, Artificial Intelligence (AI), Pandas, Deep Learning, Machine Learning Operations (MLOps)

Software Developer

2011 - 2012
Key-Soft plc
  • Participated in the development of a billing software.
  • Designed and maintained databases using PL/SQL.
  • Developed components of the billing software product.
  • Supported the development of an online bookstore in PHP, SQL.
  • This was an internship program besides the university.
Technologies: PHP, PL/SQL, C++

Software Developer

2009 - 2009
Rise FM
  • Created interactive banners for the website of the company.
  • The main development was done in Flash (ActionScript), HTML, CSS, and PHP.
  • Collected reviews and feedbacks from the viewers of the website.
  • Maintained specific parts of the website based on the reviews.
  • This was a summer job besides high school.
Technologies: PHP, CSS, HTML, Flash, Flash ActionScript

Experience

Colorgram

https://github.com/denesbartha/Colorgram
While working at the National University of Singapore one of my projects was to create a much efficient representation of the Succinct Colored de Bruijn Graph data structure used for DNA assembly, compression, bubble calling and to detect variations between individuals of a population.

Tree Graph Labeling

https://github.com/denesbartha/tree-graph-labeling
For my Master's thesis, I needed to use an efficient tree labeling algorithm. Because at the time the currently available algorithms were mostly theoretical (without any libraries), I have decided that I give a concrete solution to the problem. First I implemented my algorithm in C++ and then I gave an alternative implementation in Python.

Reconstruction of Rooted Directed Trees

https://github.com/denesbartha/RRDT
While working on my Ph.D., I was mostly concentrating on problems of Statistical Bioinformatics. One particular problem was how to reconstruct tree graph structures from given frequencies of subgraph information. I gave a concrete algorithm for the rooted directed trees' problem and published my results in a paper.

Education

2014 - 2019

Ph.D. in Computer Science

Eötvös Loránd University - Hungary

2012 - 2014

Master's Degree in Computer Science

Eötvös Loránd University - Hungary

2009 - 2012

Bachelor's Degree in Computer Science

Eötvös Loránd University - Hungary

Certifications

MAY 2017 - PRESENT

Associate Android Developer

Google

DECEMBER 2015 - PRESENT

Foundation Certificate in Software Testing

ISTQB

JUNE 2012 - PRESENT

Software Information Technologist

Eötvös Loránd University

Skills

Libraries/APIs

Scikit-learn, Pandas, NumPy, Sage, Keras, TensorFlow, XGBoost, PyTorch

Tools

BigQuery, Git, PyCharm, CLion, Sublime Text 3, Sublime Text, Flash, Maple, MATLAB, Tableau, AutoML, Cloud Dataflow, Apache Airflow, Terraform, You Only Look Once (YOLO), Microsoft Power BI

Languages

Python, C++, C, R, MicroPython, Flash ActionScript, HTML, CSS, PHP, Assembly, SQL, Java, C#, Rust, JavaScript, Dart

Platforms

Google Cloud Platform (GCP), Linux, Raspberry Pi, Ubuntu, Android, Jupyter Notebook, Amazon Web Services (AWS), Kubeflow, Kubernetes

Frameworks

Boost, Django, Android SDK

Paradigms

Scrum

Storage

PL/SQL, MySQL

Other

Artificial Intelligence (AI), Machine Learning, Machine Learning Operations (MLOps), Chatbots, Computer Vision, Deep Learning, ESP32, Large Language Models (LLMs), Hardware Development, PCB Design, BERT, Natural Language Processing (NLP), Language Models, Data Build Tool (dbt), Google Colaboratory (Colab), Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), Sensor Data, IoT Protocols

Collaboration That Works

How to Work with Toptal

Toptal matches you directly with global industry experts from our network in hours—not weeks or months.

1

Share your needs

Discuss your requirements and refine your scope in a call with a Toptal domain expert.
2

Choose your talent

Get a short list of expertly matched talent within 24 hours to review, interview, and choose from.
3

Start your risk-free talent trial

Work with your chosen talent on a trial basis for up to two weeks. Pay only if you decide to hire them.

Top talent is in high demand.

Start hiring