Roland is available for hire
Hire RolandRoland Szabo
Verified Expert in Engineering
Full-stack Developer
Location
Oradea, Romania
Toptal Member Since
September 22, 2022
Roland is a software developer with experience in various machine learning projects, from NLP and computer vision to time series data. He worked for small startups and big companies, such as Google. For the last 1.5 years, he has worked as a freelance consultant helping companies get started with machine learning projects.
Portfolio
Self-employed
AI Design, Advisory, Python, FastAPI, fastText, SpaCy...
Archive360
Python, FastAPI, Spark, Apache Kafka, Kubernetes, Docker, Azure Cosmos DB...
Laif Computation
Python, Apache Kafka, TensorFlow, Artificial Intelligence (AI), Computer Vision...
Experience
Availability
Part-time
Preferred Environment
PyCharm
The most amazing...
...thing I've developed is a SQL query to estimate how many rooms are booked but not used for meetings.
Work Experience
Machine Learning Consultant
2021 - PRESENT
Self-employed
- Designed the data annotation process for a structured information extraction project from job vacancies.
- Improved the accuracy of a logo recognition model that a team of five has been working on for months.
- Improved workflows related to ML development, specifically data versioning, experiment tracking, and code organization.
Technologies: AI Design, Advisory, Python, FastAPI, fastText, SpaCy, Hugging Face Transformers, Scikit-learn, Docker, Technology Consulting, Startup Consulting, PyTorch, OpenAI GPT-4 API, OpenAI GPT-3 API, React
Machine Learning Tech Lead
2019 - 2020
Archive360
- Created the architecture for microservices used to extract data from Office 365.
- Modernized web services to use the FastAPI framework for HTTP APIs, reducing latency and increasing throughput.
- Designed the service bus architecture for a scalable system that processes and extracts information from documents.
Technologies: Python, FastAPI, Spark, Apache Kafka, Kubernetes, Docker, Azure Cosmos DB, Artificial Intelligence (AI), Machine Learning, AI Design, Generative Pre-trained Transformers (GPT), Natural Language Processing (NLP), GPT, SQL
Machine Learning Tech Lead
2018 - 2019
Laif Computation
- Led a team of five to build a real-time surveillance camera processing system to recognize faces, identify intruders, and detect dangerous objects such as guns and knives. The models were tuned to achieve low latency.
- Designed the architecture and led a team that built a GUI system to enable non-technical people to create ML pipelines without having to write code.
- Analyzed historical data for an oil company to improve its shipping and maintenance processes.
Technologies: Python, Apache Kafka, TensorFlow, Artificial Intelligence (AI), Computer Vision, Machine Learning, AI Design, GPT, Generative Pre-trained Transformers (GPT), Natural Language Processing (NLP), Technology Consulting, Google Cloud, Chatbots
Software Engineer
2016 - 2018
- Developed a meeting room recommendation engine to speed up the process of setting up multi-office meetings.
- Analyzed events from calendars to find aggregate patterns in how people use their time.
- Developed pipelines to process and aggregate historical calendar event data in an efficient way.
Technologies: Java, Apache Beam, Artificial Intelligence (AI), Machine Learning, Natural Language Processing (NLP), GPT, Generative Pre-trained Transformers (GPT), SQL
Site Reliability Engineer
2014 - 2016
- Created machine learning models to detect anomalies on data monitored by production servers.
- Developed automation to turn up and down clusters for Gmail in new data centers.
- Built planning and forecasting modules to estimate future resource demand for Gmail.
Technologies: Python, Go
Experience
Job Advertisement Parsing
I defined the data annotation process for job advertisements, clarified the taxonomy given by domain experts and recruiters, trained machine learning models, and developed an API to use the models for parsing job advertisements.
ESG Classification Through Leader Articles
A Python-based website where a user could submit a Twitter handle. The website would scrape all the links they wrote, identified if articles were written by that user or if they mentioned them, and classified them based on an ESG-related taxonomy.
Predictive Maintenance for Factories
Developed a predictive maintenance module for forecasting when industrial machinery will require maintenance. I analyzed the data and trained anomaly detection models to identify with a seven-day lead when problems would occur.
Monitoring for Garbage Tracks
A computer vision app to analyze what is thrown into a garbage truck. When the wrong items are thrown into the truck (for example eg: biodegradable material into the recyclable truck), an alarm is raised.
Extracting Financial Information from Scanned Documents
A Python backend to run LayoutLM models to extract tables, forms, and other information from the financial document to automate their processing. The app features a UI to allow humans to correct the information and improve the ML models by performing periodic retraining.
Education
2011 - 2014
Bachelor's Degree in Computer Science
Babes Bolyai University - Cluj-Napoca, Romania
Skills
Languages
Python, SQL, Go, Java, Python 3
Libraries/APIs
Scikit-learn, PyTorch, SpaCy, TensorFlow, React
Tools
Doccano, PyCharm, Apache Beam, Celery
Other
FastAPI, fastText, Artificial Intelligence (AI), Machine Learning, AI Design, Natural Language Processing (NLP), GPT, Generative Pre-trained Transformers (GPT), Prodigy, Computer Vision, Large Language Models (LLMs), Advisory, Technology Consulting, Startup Consulting, Chatbots, Document Processing, Hugging Face Transformers, Edge Computing, Transformers, Language Models, OpenAI GPT-4 API, OpenAI GPT-3 API
Platforms
Docker, Apache Kafka, Kubernetes, Jupyter Notebook
Frameworks
Spark
Paradigms
Object-oriented Programming (OOP)
Storage
Azure Cosmos DB, Google Cloud
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