Hiring a Data Scientist is Hard Anyone who has successfully hired a data scientist will tell you it can be tough. All the others who failed at it will tell you it’s tough, too. When we started NarrativeDx, I spent six months trying to hire a data scientist. I put up all sorts of postings on various sites, tapped my own network as well as DreamIt’s network, but in the end only found our Chief Data Scientist because he was officing in the room next door to us. Sometimes you just have to get lucky.
Do You Need a Data Scientist? The value provided by good data scientists operating on rich data sets is well understood. This situation, however, isn’t as prevalent as it might seem. The best data science models are only as good as the data used to train them. When we founded the company, we wanted to hire a data scientist - but we weren’t ready yet. We thought we weren’t ready because we couldn’t afford them, but in reality, we had nothing for them to do. Until we had initiated early pilots and annotated roughly 50,000 comments by hand, there was no reliable gold standard data on which to build the models.
Fortunately, by the time we had built our gold standard by hand, we had also shown enough traction to raise funds and hire a data scientist. Before you set out to hire, make sure you understand the problems and the data you need for a data scientist to be effective.
Can You Afford a Data Scientist? Data scientists are in high demand. Google, Apple, and other Silicon Valley giants snap up young data science PhDs before they even graduate - sometime without projects for them to work on, all to stockpile talent in a data science arms race. For the rest of us out there, that means data scientists are very expensive.
According to Angel List, data scientists in Austin have an average salary of $93,000. From my personal experience, the sorts of people I’d be willing to hire as employee #1 of a healthcare AI platform like NarrativeDx may go for that salary, but would require a significant equity component. For cash-only compensation, I’ve seen top talent in Austin go for north of $150,000 - easily.
Where Can I Find a Data Scientist? Personal references have been the most effective hires in my opinion, although even those with large networks may have trouble finding people willing to leave cushy positions for the inherent risk of joining a relatively new company. I’ve found Angel List to be a good resource for finding a large volume of potential data scientists, but the quality of candidates in terms of experience and commitment is all over the map.
One trick that has helped me as I’ve gotten better at searching is to research companies that I respect and ask their data science leaders if they have anyone in their network that is looking. The success rate is a lot higher than you might expect.
Narrative Dx: The AI Platform for Patient Experience At NarrativeDx, our data scientists and engineers have built a powerful AI platform for patient experience using natural language processing and machine learning to analyze qualitative data from patient feedback and comments about their care experiences.
Over the years, we’ve seen hospitals struggle to make sense of their patient experience surveys from various data sources, which is an incredibly manual and time-consuming undertaking for their in-house specialists. This process can take weeks or months to complete in-house at a hospital or clinic, slowing down critical business decisions that should ultimately improve referrals, revenue and overall care experiences.
As an alternative to the aforementioned manual analysis of qualitative patient feedback data, the NarrativeDx AI platform automates the processing and analysis with a healthcare-specific ontology of over 5,000 categories. Our streamlined and intuitive dashboards allow hospital and patient experience leaders to hear, understand, and act on the voice of their patients thanks to the evidence we build for them from patients’ own words.
Kyle Robertson, JD, is the Founder and CEO of NarrativeDx.