Margarida Campos, Developer in Lisbon, Portugal
Margarida is available for hire
Hire Margarida

Margarida Campos

Verified Expert  in Engineering

Machine Learning Developer

Location
Lisbon, Portugal
Toptal Member Since
January 6, 2022

Margarida is a data scientist and machine learning developer with 5+ years of experience in the field. At OutSystems, she won the Top Performer Award of 2020. She has been a data science teacher since 2019, having taught hundreds of students in machine and deep learning. Margarida has also worked as a researcher, having won an international competition with the development of an original QA model in 2021. She currently works as a freelancer.

Portfolio

Instituto Superior Técnico - University of Lisbon
ChatGPT, Natural Language Processing (NLP), Deep Learning, Statistical Modeling...
Instituto Superior Técnico - University of Lisbon
Information Theory, Deep Learning, Higher Education, Education, Chatbots
Le Wagon
Python, Data Science, Data Analysis, English, Statistical Analysis...

Experience

Availability

Part-time

Preferred Environment

Python, R, GitHub, PyTorch, TensorFlow

The most amazing...

...thing I've built from scratch and with limited computational resources is an original biomedical QA system that won one of the batches of the BioASQ challenge.

Work Experience

PhD Research Fellow

2023 - PRESENT
Instituto Superior Técnico - University of Lisbon
  • Worked on a thesis entitled "Uncertainty Quantification of Language Models for Medical Applications."
  • Studied the use of conformal prediction and other statistical methods to provide confidence intervals for language models' certainty on their outputs.
  • Tested proposed techniques with ChatGPT and other LLMs on biomedical and medical datasets.
Technologies: ChatGPT, Natural Language Processing (NLP), Deep Learning, Statistical Modeling, OpenAI GPT-4 API

Invited Teacher

2022 - PRESENT
Instituto Superior Técnico - University of Lisbon
  • Taught Master's students in Information and Communication Theory and Deep Learning courses.
  • Developed Python notebooks to add a practical and hands-on component to courses.
  • Created quizzes to monitor students' progress in a pedagogical manner.
Technologies: Information Theory, Deep Learning, Higher Education, Education, Chatbots

Lead Teacher of Data Science

2020 - PRESENT
Le Wagon
  • Taught ten batches of students in data analysis, decision science, machine learning, and deep learning; 100% of students' feedback was positive.
  • Guided the implementation of dozens of students' data science projects from beginning to end, including facial recognition using CNNs, news classification, and stock prediction using RNNs.
  • Represented Le Wagon in conference Future.Works Tech Conference. Feedback for the speaker given by over 50 people was 9.5/10.
Technologies: Python, Data Science, Data Analysis, English, Statistical Analysis, Statistical Data Analysis, Artificial Intelligence (AI), Scraping, Web Scraping, APIs, Scikit-learn, Pandas, Chatbots

Researcher

2020 - 2021
LASIGE
  • Developed a model for biomedical question answering using deep learning to participate in the BioASQ challenge. The model won one of the test batches.
  • Published a paper entitled "Post-processing BioBERT and Using Voting Methods for Biomedical Question Answering," presented at the CLEF 2021 conference.
  • Wrote and defended a thesis detailing the development of the QA model and its novel methods, graded 19/20.
Technologies: PyTorch, Deep Learning, Generative Pre-trained Transformers (GPT), Natural Language Processing (NLP), GPT, English, Statistical Modeling, Statistical Analysis, Statistical Data Analysis, Artificial Intelligence (AI), Scraping, Web Scraping, APIs, Docker, Scikit-learn, Pandas, Chatbot Conversation Design, Chatbots

Data Scientist

2018 - 2020
OutSystems
  • Developed and oversaw the implementation of a lead scoring model based on demographic and website behavior data.
  • Created a spam detection model for forum posts that saved hundreds of hours of manual labeling.
  • Designed and implemented over ten Shiny dashboards for metric monitoring for sales, finance, and marketing departments. The dashboards included forecasts for revenue, sales, and conversion rates.
Technologies: R, Python, Microsoft Power BI, SQL, Snowflake, English, Statistical Modeling, Statistical Analysis, Statistical Data Analysis, Data Modeling, Scraping, Web Scraping, Analytics, APIs, Docker, Scikit-learn, Pandas

Biomedical Question Answering System

https://github.com/lasigeBioTM/BioASQ9B
A deep learning-based model for biomedical QA for exact answer retrieval.

The model used transfer learning on different datasets and novel post-processing techniques, and it was implemented using PyTorch.

Languages

Python, R, SQL, Snowflake, JavaScript, HTML5, CSS

Libraries/APIs

Scikit-learn, Pandas, PyTorch, TensorFlow

Paradigms

Data Science

Other

Statistics, Probability Theory, Machine Learning, Deep Learning, Natural Language Processing (NLP), Data Visualization, Data Analysis, English, Statistical Modeling, Statistical Analysis, Statistical Data Analysis, Analytics, GPT, Generative Pre-trained Transformers (GPT), Predictive Modeling, Programming, Data Modeling, Time Series Analysis, Artificial Intelligence (AI), Scraping, Web Scraping, APIs, Customer Segmentation, Chatbot Conversation Design, OpenAI GPT-4 API, Chatbots, Algorithms, Reinforcement Learning, Bayesian Statistics, Information Theory, Higher Education, Education

Tools

GitHub, ChatGPT, Microsoft Power BI

Platforms

Docker

2019 - 2021

Master's Degree in Data Science and Engineering

Instituto Superior Técnico, University of Lisbon - Lisbon, Portugal

2012 - 2017

Bachelor's Degree in Applied Mathematics and Computation

Instituto Superior Técnico, University of Lisbon - Lisbon, Portugal

MARCH 2019 - PRESENT

Neural Networks and Deep Learning

Coursera

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