About
I am a data scientist and machine learning engineer with 5+ years of experience building models for computer vision, natural language processing, and tabular datasets. With a B.S. in neuroscience and an M.S. in computer science, I am especially excited about the development of neural networks and the increasing complexity of problems they can solve. Over the past several years, I have had the opportunity to work on analytics for large-scale medication adherence outreach programs, multi-task neural networks for brain microscopy image segmentation, transformer-based NLP models to detect COVID-19 outbreaks from news articles, explainable machine learning for fraud detection, and more.
Currently, I am working at Superlinear as a Generative AI Team Lead and Solution Architect.
Data Scientist & Machine Learning Engineer
Experienced in all aspects of the data science pipeline, including data exploration, feature engineering, model training, and model deployment.
- University: Georgia Tech, UMiami
- Employer: Radix
- Location: Brussels, Belgium
- Degree: B.S., M.S.
- Email: j.miano@outlook.com
- Languages: English, French, Spanish
Skills
Programming Languages
- Python
- SQL
- MATLAB
- Java
- C
ML & Big Data
- PyTorch
- Transformers
- Scikit-learn
- PySpark
- Dask
Visualization
- Matplotlib
- Seaborn
- Plotly
- Streamlit
- Tableau
Techniques
- Deep Learning
- Feature Engineering
- Ensemble Methods
- Unsupervised Learning
- Prompt Engineering
Data Domains
- Computer Vision
- Natural Language Processing
- Audio & Speech
- Tabular Datasets
- Time Series
DevOps & Cloud
- Docker
- Git
- Azure
- GCP
- AWS
Resume
Click here to view my resume as a PDF.
Summary
Joseph Miano
Data scientist and machine learning engineer with 5+ years of experience in computer vision, natural language processing, tabular datasets, and deep learning.
- Brussels, Belgium
- j.miano@outlook.com
Education
Master of Science in Computer Science
Aug 2020 - Dec 2021
Georgia Institute of Technology, Atlanta, GA
- Machine Learning Specialization
- Graduate Research Assistant at the Georgia Tech Research Institute
- Coursework in deep learning, computer vision, natural language processing, and machine learning theory
Bachelor of Science in Computer Science
May 2018 - May 2020
Georgia Institute of Technology, Atlanta, GA
- 2nd B.S.
- Coursework in computer science and mathematics
- Specializations in theory and artificial intelligence
Bachelor of Science in Neuroscience
Aug 2012 - May 2016
University of Miami, Coral Gables, FL
- Minors in Finance and Chemistry
- Research in cellular neuroscience
- Pre-medical track with medical shadowing experience
Internships
AI & Machine Learning Summer Associate
Jun 2021 - Aug 2021
JPMorgan Chase & Co., Remote, USA
- Developed object-oriented Python code to enable explainability and interpretability of credit risk assessment models
- Presented results and conclusions to the broader intern group and organization (20+ colleagues)
Software Engineering Summer Intern
Jun 2019 - Aug 2019
American Express, Phoenix, AZ
- Trained natural language processing machine learning models using Python to automate incident ticket routing
- Explained summer project and results to VP-level organization (40+ colleagues) during end-of-internship presentation
Work Experience
Senior Machine Learning Engineer → Team Lead
Mar 2023 - Present
Radix, Brussels, Belgium
- Lead a team of 7 machine learning engineers, which includes career growth mentorship, organizational planning, and project delivery support
- Spearheaded a project to improve the efficiency of a medicine production pipeline via machine learning in collaboration with a pharmaceutical partner
- Developed deep learning models (convolutional autoencoders) to denoise barcode images for a project with a retail partner, leading to a 15%+ improvement in barcode recognition accuracy
- Modeled (predictive) patient outcomes after surgical interventions for a project with a medical partner, facilitating patient expectation management at the various stages of their journey
AI & Machine Learning Senior Associate
Feb 2022 - Mar 2023
JPMorgan Chase & Co., New York, NY
- Engineered 100+ features for customer authentication risk assessment models, specifically to mitigate digital authentication risk
- Trained machine learning models to predict fraudulent customer authentication events, balancing customer service experience (i.e., false positives) with fraud risk (i.e., false negatives)
- Coordinated the explainable AI track for the inaugural 2022 JPMorgan Chase AI Summit, which brought together speakers from across the firm to present on how model explainability methods are being applied
Graduate Research Assistant
Sep 2020 - Dec 2021
Georgia Tech Research Institute, Remote, USA
- Implemented neural natural language processing models (BERT and RoBERTa) to automate COVID-19 outbreak detection using web-scraped news article contents
- Published a paper as first author in the Springer Lecture Notes in Artificial Intelligence as part of the 2021 Artificial Intelligence in Medicine Conference
Research Assistant
Aug 2018 - Jul 2020
Neural Data Science Lab @ Georgia Tech, Atlanta, GA
- Developed multi-task convolutional neural network for segmentation and classification of mouse brain x-ray microtomography data
- Presented joint poster at the Allen Institute BioImage Informatics 2019 Conference (funded with PURA Travel Award)
- Collaborated to publish 3 papers (linked in the Papers section)
Consultant → Senior Consultant
Aug 2016 - Apr 2018
CVS Health, Woonsocket, RI
- Identified patients at risk of medication non-adherence in outcomes-based contracts and executed adherence outreach programs
- Quality-tested 50+ features for an enterprise-level predictive modeling project in collaboration with stakeholders from several departments
- Coordinated onboarding for 8 new hires and guided curriculum development of the onboarding program, including the addition of a new SQL training
Projects
Hover or click on the images below to get a summary and link for each project.
In this project, I deployed a random forest model and dashboard on AWS visualizing data and predictions for diabetes hospital readmissions.
In addition to interactive visualizations, the dashboard enables the user to upload their own data and download model predictions.
Of the various models trained and tested, random forest performed the best, and the two most important features predicting hospital readmission were the number of lab procedures and the number of medications for the patient.
In this group project, we studied the impact of noisy samples and pruning neural networks on image and audio through the lens of the cognitive science model of graceful degradation.
My focus in the project was the audio data, for which I trained 1D convolutional neural networks to process raw audio and 2D ones to process spectrogram-transformed audio.
We found that our neural networks were quite resilient to pruning when retrained and could learn to adapt to noisy inputs.
In this group project, we studied the relationship between medication review text, metadata, and review usefulness.
My focus in the project was exploratory data analysis and training of text-only DistilBERT models to process the text and hybrid DistilBERT models to process the text and metadata jointly.
Overall, we were able to predict review usefulness successfully from both the text only and the metadata only, but that the hybrid model performed best.
In this group project, we developed a prototype machine learning inference system that leverages pruning of MTL (multi-task learning) neural networks.
My focus in this project was the multi-task neural network architecture design and implementation, as well as the experiments related to pruning and varying task-head length.
We found that pruned and fine-tuned MTL neural networks achieved higher accuracy-latency trade-offs than single-task models.
In this group project, we developed a framework for interactive story generation by leveraging GPT-2.
My focus in this project was to fine-tune GPT-2 to enable prompt-based story generation and to develop an interface for users to interact with.
By breaking up the story generation process into smaller chunks, we were able to create a compelling user experience for user-driven stories.
Papers
While studying at the Georgia Institute of Technology, I had the opportunity to contribute to 4 published papers and complete a thesis.
Using Event-Based Web-Scraping Methods and Bidirectional Transformers to Characterize COVID-19 Outbreaks in Food Production and Retail Settings
1st Author | 2021
A three-dimensional thalamocortical dataset for characterizing brain heterogeneity
4th Author | 2020
Multi-task learning for neural image classification and segmentation using a 3D/2D contextual U-Net model
Thesis | 1st Author | 2020