About Me

Eamon Fleming

Data scientist with background in healthcare research, writing, and design. Fascinated by the creativity in math and compelled to beauty in presentation, I enjoy solving problems by distilling powerful datasets into elegant returns.

After two years studying economics and statistics at Harvard University, I decided to chase my growing interest in medical research and pursue a degree in neurobiology. Since graduation, I've built out a career in healthcare research in the Boston area.

This has included a year with a Parkinson's Disease genetics research lab; two years at a small startup developing a non-invasive, digital respiratory monitor for surgical contexts, and three years with a nephrology research team at Tufts Medical Center.

As I came to a natural endpoint in my research projects in the fall of 2018, I was increasingly fascinated by the data analytics component of my work. I decided to do something I had been considering for years: expand my analytics and modeling toolset, and transition toward a full-time career in data science.

I relocated to Austin, and gained admission to the Data Science Immersive course at General Assembly. This intensive, four-month long program gave me a strong foundation in Python, Pandas, machine learning, big data, underlying statistics, and the major toolkits, compute resources, and procedures that define the field.

After completing the GA program, I joined a small education tech startup focused on building a reading and vocabulary-building app for non-native English speakers. As the company's primary data scientist, I took on a diverse set of responsibilities:

  • Built from scratch and iteratively managed data assets underpinning the app. This included extensive data cleaning; Python scripting in Jupyter notebooks; optimizing data structures; designing data science and mathematical models to create desired features; integration from multiple sources (SQL, JSON, scraping, raw text); output to multiple sources (Hive, JSON, data warehouse) in specific formats.
  • Contributed to development and conceptual design of a sophisticated proprietary lemmatizer for processing textual data. Leveraged large language model technology to help automate decisions output by the lemmatizer.
  • Hired and supervised a rotating team of interns, delegating and managing components of projects ranging from exploring and cleaning novel data sources to designing web applets for data review.
  • Significant latitude and responsibility in terms of project design and direction. Reported directly to the CEO and coordinated as necessary with DevOps.

Skills Developed:  Python · NLP · SQL · Pandas · Jupyter · Regex · Data Modeling · JSON · XML · VSCode · ETL · AWS Cloud · LLMs · Web Scraping · Versioning · Team Management

As I've gotten into the field, I've tried to stay open-minded and pursue a diverse set of interests. I'm not closed off to any project or application, as long as I'm strengthening my skillset. That said, certain areas have definitely caught my attention. In the short term, two of the areas I'm most interested in are:

  • Natural Language Processing: As a writer and lover of language, this area fascinates me. The last two years have seen enormous advancements, with the release of Elmo by the Allen Institute, Google's BERT, and OpenAI's GPT-2, so the time is ripe. This area also seems critical to the expressiveness of higher machine intelligence as time goes on.
  • Healthcare Applications: Whether it be automating imaging analyses, finding patterns in massive patient databases, aiding in protein folding analysis or drug discovery, or churning through research papers to try and predict new directions, the healthcare field is perhaps the greatest largely untapped data reservoir that we have. As data science in healthcare finally picks up steam, it's very tempting to return to the field in which I have the most experience and find new ways to help people.

In a broader sense, I'm also very interested in the long term progression toward general artificial intelligence (GAI). I would love to work with a team focused on explainability in black box modeling techniques like neural networks, as I think this will have great implications toward GAI, as well as obvious short term benefits. I'm also very interested in innovating the design of neural network architectures to allow organization that more closely competes with the incredible organizational complexity of the human brain.

Thanks for visiting!

Feel free to reach out with any questions or inquiries about my work, opportunities, or whatever else.

Contact me!