Dragos Dima, Developer in Bucharest, Romania
Dragos is available for hire
Hire Dragos

Dragos Dima

Verified Expert  in Engineering

Machine Learning Engineer and Developer

Location
Bucharest, Romania
Toptal Member Since
October 20, 2022

Dragos is a passionate machine learning engineer with five years of experience in artificial intelligence. He is well-grounded in natural language processing, Python, and SQL. Dragos has an excellent knowledge of deep learning frameworks such as TensorFlow and PyTorch.

Portfolio

Microsoft
Natural Language Processing (NLP), Generative Pre-trained Transformers (GPT)...
Raiffeisen Bank Romania
Python 3, TensorFlow, OCR, GPT, Generative Pre-trained Transformers (GPT)...

Experience

Availability

Part-time

Preferred Environment

Linux, Python 3, TensorFlow, PyTorch, SQL, Pandas, Jupyter Notebook, PyCharm

The most amazing...

...software I've developed is a text-to-text approach for detecting and generating keywords on small texts.

Work Experience

Machine Learning Engineer

2021 - PRESENT
Microsoft
  • Developed an intent classification model for a web page using a multi-modal approach.
  • Built a sentence importance ranking based on an attention mechanism and a transformer encoder.
  • Developed a multilingual question generator using state-of-the-art MetaTrader 5.
Technologies: Natural Language Processing (NLP), Generative Pre-trained Transformers (GPT), GPT, Computer Vision, Linux, Machine Learning, Deep Learning, Deep Neural Networks, Transformer Models, SQL, PyTorch, Data Analysis, Labeling, OpenAI GPT-3 API, Pandas, Python, Artificial Intelligence (AI), Named-entity Recognition (NER)

Machine Learning Engineer

2019 - 2021
Raiffeisen Bank Romania
  • Created a tool for clustering and finding patterns in the market sector clients.
  • Developed a tool for entity recognition on scanned documents using state-of-the-art natural language processing techniques such as contextualized word embeddings.
  • Built a churn predictive model based on client behavior using transactional data.
Technologies: Python 3, TensorFlow, OCR, GPT, Generative Pre-trained Transformers (GPT), Natural Language Processing (NLP), Machine Learning, Computer Vision, Deep Learning, Deep Neural Networks, Data Structures, SQL, Data Science, Data Analysis, Pandas, Python, Artificial Intelligence (AI), Web Scraping, Time Series, Predictive Analytics, Amazon Web Services (AWS)

Keywords Generation

I created a keywords generation tool based on the transformer architecture. The encoder was loaded with a pre-trained transformer encoder. It was trained on multiple public datasets and could generate both present and absent keywords from texts.

Information Extraction From PDF

Developed a tool for entity recognition on scanned documents using state-of-the-art Natural Language Processing techniques such as contextualized word embeddings, recurrent neural networks, and conditional random fields in TensorFlow.

WebPage Classification

Developed a multi-modal deep-learning approach for webpage classification using PyTorch. Among the architectures used in the final product, the most notable was an ensemble of Vision Transformer, Roberta, and Layout LM v3.

Table Extraction/OCR Tool

A PyTorch-based multi-modal architecture for extracting tables and other information (with formats) from PDFs. The architecture was based on Swin-Transformer for the encoder and BART for the decoder. I was involved in pre-training and fine-tuning for the task in question.
2019 - 2021

Master's Degree in Artificial Intelligence

Politehnica University of Bucharest - Bucharest, Romania

2015 - 2019

Bachelor's Degree in Computer Science

Politehnica University of Bucharest - Bucharest, Romania

JUNE 2021 - JUNE 2024

TensorFlow Developer Certificate

Google

Libraries/APIs

TensorFlow, PyTorch, Pandas

Tools

Named-entity Recognition (NER), PyCharm

Languages

Python 3, Python, SQL

Paradigms

Data Science, Object-oriented Programming (OOP)

Platforms

Amazon Web Services (AWS), Linux, Jupyter Notebook

Other

Machine Learning, Neural Networks, Deep Learning, Natural Language Processing (NLP), Artificial Intelligence (AI), Transformer Models, GPT, Generative Pre-trained Transformers (GPT), Generative Pre-trained Transformer 3 (GPT-3), Labeling, Predictive Analytics, Computer Vision, Data Structures, Data Analysis, OpenAI GPT-3 API, Web Scraping, Data Scraping, Time Series, OpenAI GPT-4 API, Information Retrieval, OCR, Deep Neural Networks, Convolutional Neural Networks (CNN)

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