Highlights from a panel discussion on the promise — and responsibility — of data-powered innovation in healthcare, finance, and smart cities.
From healthcare and finance to public safety and government, organizations are finding new ways to solve business problems with non-traditional data sources. During NYC Open Data Week, Luminary Labs hosted a panel discussion that explored the use of datasets from unexpected sources in the pursuit of innovation.
Carlos Rodarte is founder and managing director of Volar Health, where he helps innovative health companies refine their product strategy and rapidly commercialize their solutions. He serves on the editorial board of the Digital Biomarkers journal; founded HealthRhythms, a AI-driven digital health startup that utilizes smartphone sensors to better understand human behavior; and previously oversaw strategic partnerships at PatientsLikeMe.
Jeremy Baksht is global head of alternative data at Bloomberg, where he builds cutting-edge financial data products that draw on diverse data sources — including sentiment, transactions, satellites, and emails. Jeremy has held leadership roles in both traditional financial institutions and fintech startups; co-founded his own data startup, Qineqt; and is an active angel investor and blockchain enthusiast.
Mitsue Iwata is project manager at the NYC Mayor’s Office of Data Analytics, where she leads the City’s Citizens Connect Program with Waze, a data sharing program that yields actionable insights to improve public safety. She also runs Data Drills with the Office of Emergency Management, which takes a data-based approach to preparing for natural disaster response.
In finance, health, and public safety, virtualizations and simulations of real-world patterns can transform the way we run businesses, treat patients, and respond to emergencies.
New York City is exploring Waze data to analyze traffic conditions and improve public messaging. The city’s open data law, enacted in 2012, has prompted a shift in mindset among stakeholders. Mitsue acknowledged that as the city’s use of data has become more open and transparent, people are seeing the value and the quality of information has improved.
Finance is also seeing a shift in the use of data. Investors have always used traditional data sources, but the availability of new data is enhancing the ability to make decisions. How quickly is a company like Airbnb changing an industry like hospitality? Jeremy said the modern mosaic of data sources can help us quantitatively answer questions in new ways.
In an age where all data can be health data, there’s “so much cool stuff emerging,” Carlos said. Breath can be a diagnostic for lung cancer, voice can be a diagnostic for Parkinson’s, and an Apple Watch can detect diabetes or the symptoms of kidney failure. Carlos thinks of the human body as an API that’s emanating data — and we have barely begun to develop tools to capture and analyze the information.
It’s easy to feel hopeful, but new uses of data are not without risk. A significant portion of the discussion revolved around the responsible use of data.
Carlos urged organizations to view end users as collaborators, and to be as open and transparent as possible. Bring consumers, patients, or citizens into the process for honest conversations, and be prepared to explain what data is captured and how it’s used.
Jeremy also spoke in favor of verifying the provenance of all data. How and where did your organization acquire it? Is it encrypted? Does it include personally identifiable information? How do you know the terms of service weren’t breached? If you’re leading a data initiative, confident answers to these questions are table stakes.
Mitsue said New York City’s ethical algorithms task force is prompting conversations about the equitable and accountable use of data, and Sara pointed out that Bloomberg’s Data for Good Exchange and DJ Patil’s code of ethics for data scientists are important steps toward mitigating risk.
But not everyone in the room was directly involved in the collection or analysis of data, which prompted an attendee to ask: What is the responsibility of the average citizen or “non-data” professional? The panel offered several recommendations:
Ask questions. You don’t have to be a coder or engineer to work in tech, and you don’t need to be a data scientist to ask where the data comes from, who’s doing the analyzing, and which problems are being solved. Be persistent and curious, and don’t be afraid to dig deeper.
Think like a data anthropologist. “I’m not a data scientist, but I consider myself a ‘data anthropologist,’” Carlos said. He seeks to understand the context in which data is collected, and said it’s dangerous to make assumptions about the quality or specificity of data without interrogating it. Just like an anthropologist looks at history critically, we need to be critical of the socio-economic factors that influence data collection and analysis.
Bet on imperfection. There is one assumption that’s safe to make: Mitsue cautioned that clean and perfect datasets are very rare. Something is always interpolated, changed, or parsed out — acknowledge assumptions, and adjust accordingly when it comes time to make decisions based on that data.
Forge partnerships. Brilliant statisticians and data scientists don’t typically have domain expertise. If you have industry knowledge and a deep understanding of the business problems to be solved, Jeremy recommends collaborating with data experts and building relationships where you can learn from each other. There’s too much of a distinction between “data people” and everyone else, Mitsue said. Carlos agreed, and recommended a book called “The Fuzzy and the Techie,” which explores the role of the liberal arts in humanizing technology.
Photo by Denys Nevozhai on Unsplash.