Women changing the world with data science

Women changing the world with data scienceStephanie MariBlockedUnblockFollowFollowingMar 8Insight Fellows collaborating on Day 1 of the session.

As part of a 7-week program designed to help transition to a thriving career in data, Insight Fellows spend four weeks ideating, developing, and refining a project that solves a real-world problem.

Insight is committed to helping more women launch careers in tech, so in celebration of International Women’s Day 2019, we wanted to share a few of the projects completed by women participating in the January 2019 Data Science and Health Data Science sessions.

Below, we describe ten projects from our current cohort that we think represent some of the many ways that data science can make an impact and change the world.

L to R: Insight Fellows Tayler Sale, Cristiana Principato, Susanna McDonald, Hannah Lee.

Fighting human trafficking around the globeAccording to the 2018 UNODC Global Report on Trafficking in Persons, women and girls are especially vulnerable to being trafficked and make up 70% of detected victims worldwide.

A team of four Fellows from our Silicon Valley programs — Tayler Sale (Data Science), Cristiana Principato (Health Data Science), Susanna McDonald (Data Engineering), and Hannah Lee (Data Product Management) — collaborated to develop Fievel, which uses machine learning to identify online content that may be related to commercial sex trafficking of adults and minors.

Saving lives through organ donationTen percent of the world’s population is affected by chronic kidney disease, and millions die each year because of inadequate access to treatments like dialysis and organ transplant.

Unlike some other vital organs, kidneys can be transplanted from live donors — but people might be reluctant to donate; some kidney donors experience side effects like chronic fatigue.

Eva Fast (Health Data Science, Boston) built KidneyPredict, which uses medical history and demographic factors to provide personalized quality-of-life forecasts for living organ donors.

Better products for people of colorMany women around the world wear makeup, but those with darker skin tones are underserved by the cosmetic industry because most products are optimized for lighter skin, as with the potency of pigments in blush or eyeshadow, or the ranges of colors offered in foundation.

Jordanna Saeta (Data Science, Silicon Valley) created SephorAll, a tool that analyzes product reviews to provide beauty recommendations for all skin tones.

L to R: Insight Fellows Eva Fast, Jordana Saeta, Keesha Erickson, Liangliang Zhang, Rachel Mak-McCully.

Helping doctors improve their careA typical doctor-patient consultation is shorter than 15 minutes — which is not a lot of time to arrive at a correct diagnosis, especially when similar symptoms can be associated with different diseases.

Keesha Erickson (Health Data Science, Silicon Valley) analyzed text from clinical conversations and hospital reports to create Hynt, a tool that provides real-time, AI-powered recommendations to help physicians reach a correct diagnosis in less time.

Predicting risk of contracting sepsisSepsis is a condition that arises when the body’s response to an infection injures its own tissues and organs.

Each year, an estimated 30 million people develop sepsis; for around 6 million of them, the consequence is death.

However, sepsis is treatable with early intervention.

Liangliang Zhang (Health Data Science, Boston) created SepsisAlert to help healthcare providers know, in real-time, if a patient is likely to be developing sepsis, enabling them to administer life-saving antibiotics.

Diagnosing harmful sleep disordersSleep apnea is a disorder in which breathing repeatedly stops and starts during the night.

The condition is common, affecting an estimated 1 billion people across the globe — but if left untreated, it can lead to high blood pressure, heart disease, and stroke, as well as consequences related to fatigue (like car accidents caused by falling asleep at the wheel).

To help facilitate diagnosis and treatment, Rachel Mak-McCully (Health Data Science, Silicon Valley) developed Tired!.to predict increased risk of sleep apnea using data generated by wearable devices.

L to R: Insight Fellows Stephanie Gervasi, Alexandra Mably, Priscilla Addison, Zhaleh Safikhani.

Better detection of Lyme diseaseLyme disease can be transmitted through the bite of an infected deer tick, and has been reported in rural areas of Asia and parts of Europe, and is especially common in the Northeast, mid-Atlantic, and upper Midwest regions of the United States.

Each year, in the US alone, it is estimated that 300,000 people are infected, but it can be difficult to diagnose: the bull’s-eye shaped rash closely associated with Lyme disease doesn’t appear in all cases.

Stephanie Gervasi (Health Data Science, Boston) created Lyme Spotter, a tool to help doctors quickly identify patients with a high probability of testing positive for Lyme so they can facilitate immediate treatment.

Demystifying women’s clothing sizesWhereas many men’s garments are sized universally (for example, pants are labeled with measurements like waist size and inseam length), women’s clothing sizes can vary widely from brand-to-brand, making it difficult and time-consuming to find clothes that fit.

Alexandra Mably (Data Science, Remote) worked on a project called Fit Matters, where she contributed to a tool that uses transactional and customer survey data to more accurately predict if an item of clothing will fit.

Improving access to flu vaccinesWorldwide, seasonal flu epidemics result in an estimated 3–5 million cases of severe illness per year, leading to 290,000–650,000 deaths.

Vaccines can prevent or lessen effects for individuals, and help provide ‘herd immunity’ for immunocompromised individuals or those unable to receive the vaccine.

Priscilla Addison (Data Science, Seattle) built VACcess to estimate population access to the flu vaccine and identify vulnerable locations that need more providers.

Bringing equality to the job marketIt’s been repeatedly demonstrated that racial prejudices can lead employers to reject candidates based on name alone.

Anonymous processes for reviewing applicant resumes could help reduce employment discrimination and improve organizational diversity.

To help reduce bias in candidate selection, Zhaleh Safikhani (Data Science, Toronto) built Resume Anonymizer, which censors resume text indicative of applicants’ race, ethnicity, and gender, putting all applicants on a level playing field.

Are you interested in working on high-impact projects and transitioning to a career in data?.Learn more about the Insight Fellows programs and start your application today.

If you are a woman applying for Insight, and a Women Who Code member, consider applying for a scholarship!Do you want to partner with Insight in our efforts to help more women launch careers in tech?.We’d love to hear from you: partnerships@insightdatascience.



. More details

Leave a Reply