About Me
Hello, my name is Eric and I am currently a Researcher at the Data Ethics Working Group specialising on Foundation Models’ regulation at The Alan Turing Institute, GBR with the results forthcoming in the Harvard Data Science Review special issue 4. Formerly, I have been a Data Scientist at Launchpad.AI, leading a team of 4 people working on computer vision and deep neural networks and Data Research Analyst working with telemedical data for the Paracelsus Medical University of Salzburg and University of Salzburg, AUT. Before that I undertook Causality Research for Dr. Neil LLoyd at Warwick Department of Economics.
See my current passion project here: Truthfulness of LLMs. For my short introduction on cool and useful research papers in your everyday life please see my LinkedIn.
I have previously organised the AI in Healthcare Workshop at the University of Warwick and we are currently writing up the conference paper.
Overall, my Don Lavoie Fellowship at the Mercatus Centre, USA in social sciences and public policy and other work, research, and volunteering experiences at a number of organisations have enabled my teamwork and leadership skills.
I am most skilled in: Python | fast.ai, Transformers, and PyTorch, R, Stata, QGIS, Convolutional Neural Networks, Time Series Analysis, Computer Vision and Spatial Analysis.
Projects
Computer vision change detection
https://github.com/FellowshipComputerVision/Change_Detection_Fellowship_ProjectResearch Fellowship in Data Science & Machine Learning working on computer vision change detection with satellite imagery data for the Hawaii Land Trust.
I was leading a team of 4 as Scrum Master for a project of computer vision for the Hawaii Land Trust and implemented a deep convolutional neural network for satellite image recognition of land-use change detection. Moreover, I presented a research paper and submitted weekly research paper reports for the reading group. Further, I co-ordinated my project’s weekly demo day presentations by collating information, creating slides, presenting and clarifying issues to senior management. My main learnings are: computer vision, change detection using satellite imagery data, convolutional neural networks
A novel approach of mining consumer diesel prices to create more consumer transparency and oversight of price impacts.
I started this project to contribute to a new data gathering methodology for consumer oil indexes and work towards increased consumer price transparency by collecting Austrian diesel prices real-time using Python and Google Maps API as well as the Austrian regulator’s E-Control website. My main learnings are: API accessing and webscraping data, compiling a time-series index, spatial analysis
A field work research exchange undertaken at ITAM, Mexico City & University of Warwick, UK combining the fields of economics, sociology, and law to analyse the interaction of NAALC, the NAFTA labour side agreement.
The project contributed to the interdisciplinary research of human rights and macroeconomics by undertaking an extensive Qualitative Research. I focussed on NAALC, NAFTA’s labour side agreement by uniting the disciplines of sociology for migration, macreconomics for free trade, and law for labour rights. The results were presented at: International Conference for Undergraduate Research, Coventry, UK, British Conference for Undergraduate Research, Leeds, UK, and at the Undergraduate Research Support Scheme, Coventry, UK. I am grateful for the research funding granted by the Undergraduate Research Support Scheme by the University of Warwick to undertake the research exchange to ITAM, Mexico City. Many thanks go to my supervisors for their guidance and expertise: Assoc. Prof. Dr. Stefania Paredes Fuentes, University of Warwick and Dr. Christina Wagner Faegri, ITAM. My main learnings are: free trade and labour law within NAFTA, Central/North American migration
Launchpad.AI's admission test of sentiment analysis by using a pre-trained DistilBERT model to inference on IMDB dataset.
Tackling the sentiment analysis challenge, I used a pre-trained DistilBERT, a more efficient, smaller but almost as accurate version of BERT by Hugging Face designed for word completion and applied it to sentiment analysis. My main learnings are: NLP, sentiment analysis, Hugging Face plattform, pre-trained DistilBERT model
Experience
Advancing telemedicine for a better medical system in German speaking (DACH) countries by analysing needs and wants from medical experts and laypeople.
I am jointly establishing The Alan Turing Institute’s position on Foundation Models’ regulation in a team of 30 AI and industry experts, forthcoming as a co-author in the Harvard Data Science Review.
University of Salzburg and Paracelsus Medical University of Salzburg
Quantitative Data Analyst
Jun 2021 - Oct 2023
Advancing telemedicine for a better medical system in German speaking (DACH) countries by analysing needs and wants from medical experts and laypeople.
I am involved in a 4 part process. I undertook experts interview surveys for over 40 medical experts in the DACH countries. Designed balanced and weighted online surveys of 5,500 people in DACH countries. Collated, cleaned, and analysed data about attituded to telemedicine in R. Drafted final report of mixed methods study in English and German in a team of 4. My main learnings are: (qualitative) data gathering of large-scale medical and technology surveys and expert interviews and their data analysis in R.
Protecting biodiversity via land-use change detection for the Hawaii Land Trust using deep learning convolutional neural networks managing a team of 4.
The Hawaii Land Trust (HILT) wants to monitor its land for land-use land-cover changes through satellite imagery. I lead a team of 4 people training a Siamese CNN and fine-tuning it with daily satellite imagery data from Hawaii to apply it to HILT’s local environment. My main learnings are: implementing in a team computer vision change detection with deep neural networks.
Fellows join a network of Mercatus students, alumni, faculty, and scholars who are conducting and engaging with cutting edge research in contemporary political economy.
Don Lavoie Fellowship is a competitive, renewable, and online fellowship program for advanced undergraduates, recent graduates, and early-stage graduate students. Fellowships are open to students from any discipline who are interested in studying key ideas in political economy and learning how to utilize these ideas in academic and policy research. Throughout the fellowship I developed policy solutions with field-experts and international fellows to current social, political, and economic issues. My main learnings are: developing policy-making solutions to political, economic, and social challenges.
University of Warwick, Department of Economics
Research Assistant
Jul 2022 - Dec 2022
Unemployment/Self-Employment South Africa
Evolving understanding of high unemployment and low self-employment puzzle in South Africa through an historical lens on spatial mining data under Apartheid South Africa.
I was involved in 3 stage process in working with historical mining data. First, Identified microeconometric instrumental variable strategies for explaining contemporary high unemployment and low self-emplyoment puzzle. Then, I visualised innovatively spatial data and maps using QGIS and R. Finally, I applied econometric spatial inference South African historical spatial mining data utilising GitHub for version control and R for first-stage regressions in 2 least squares regressions. My main learnings are: spatial data analysis in R, visualisation of spatial data using R and QGIS, version control using Github, instrumental variable identification strategy.
Conference Presentations
Title: AI in Healthcare: A closer look at Interdisciplinarity for Improved Innovation in Future Clinical and Work Practices.
Title: How are we moving? - a new data gathering approach for consumer oil
Title: Free trade, labour rights, and migration
10 Year Anniversary of Economy for the Common Good (ECG)
Salzburg, AT
10-Oct-2020
10 year anniversary conference
Title: Work for pluralism in economics with Rethinking Economics International
Title: Macroeconomics and human rights
Education
Warwick University and year abroad at University of Salzburg
BSc (Hons) Economics
2018 - 2022
In 1964, founder J.R. “Dick” Sargent created a Department with the first few professors and graduate students. He envisioned a department that would use, and train its students to use, the critical mathematical tools needed for analysis of complex economic issues – a legacy that continues to the present day.
During my time at Warwick and a year abroad at the University of Salzburg I learnt most of my key skills that have I have taken through my career such as teamwork and working to tight deadlines. My courses most relevant to my work included amongst others neural networks, machine learning, advanced macroeconometrics, quality of social science data, time-series econometrics, microeconometrics, mathematical statistics, and game theory. I spent my free time establishing and leading the student society Rethink Economics Warwick of Rethinking Economics International on roles such as External Relations Officer and Co-ordinator organising many academic talks and panel discussions.
Grants & Affiliations
A Little More About Me
Alongside my interests in AI and machine learning some of my other interests and hobbies are:
- Volunteering at an Ukrainian Orphanage
- Writing Poetry in German, English, Spanish, and French
- Photography
- Sailing
- Sewing