Biggest Challenges Faced By A Data Scientist At Cohere Health
Shailja's biggest challenges as a Data Scientist at Cohere Health involve "data wrangling," transforming messy healthcare data like "doctors' notes that are literally scribbled," into usable formats. A second key challenge is effectively communicating complex technical findings and limitations to non-technical stakeholders.
Data Wrangling, Data Analysis, Communication, Problem-Solving, Healthcare Data
Advizer Information
Name
Job Title
Company
Undergrad
Grad Programs
Majors
Industries
Job Functions
Traits
Shailja Somani
Data Scientist
Cohere Health
Johns Hopkins University, 2020
Currently pursuing my MS in Applied Data Science at the University of San Diego (part-time online while working full-time)
Psychology
Technology
Data and Analytics
Greek Life Member, LGBTQ
Video Highlights
1. Data cleaning and wrangling is a significant part of the job, requiring skills in handling imperfect and diverse data formats. This includes working with unstructured data like handwritten doctor's notes.
2. Translating complex technical findings into easily understandable explanations for non-technical stakeholders is crucial for effective communication and collaboration.
3. Healthcare data presents unique challenges due to its varied formats, inconsistencies, and the need to handle sensitive patient information responsibly and ethically
Transcript
What is your biggest challenge in your current role?
I think I would say two things. First, data is never perfect. There are always issues, and there are always things that we might want that we might not have access to.
Sometimes data is in vastly different formats than we expect. For example, I work in healthcare, so you get doctors' notes that are literally scribbled or just crazy chicken scratch. Or maybe they type things out in abbreviations that only their office understands. So there's a lot of data wrangling and figuring all of that out.
The second thing I would say is a little bit what I answered to an earlier question, but it's translating from technical work to non-technical stakeholders. Sometimes it's hard to explain what exactly we're working on. This includes explaining limitations and even, on the opposite side, explaining insights and things that we've understood from the data.
It's challenging to explain this to people who don't necessarily understand what's in the data, what it looks like, or how it might be able to specifically generate certain insights, or what algorithms we use to generate those.