Generally speaking, I work with mobile devices to make people healthier and happier.

I conduct research on user engagement, data collection, and health behavior using mobile health apps. With what I learn, I design products/ systems to assess people's health and health behavior and ultimately help them make healthier choices. I do this as the founder of Curiosity Health, Senior Researcher in Residence at Cornell Tech, and Assistant Professor at Weill Cornell Medicine.

My background is mostly in product development and software engineering, nearly all in health and life sciences. I’ve built products, started companies, and advised even more companies. Education is genetics and computer science undergrad, human-computer interaction masters, and information science PhD.

Below is some info on what I am currently or have recently been working on. If you're curious or bored (ok, really bored), wade down into the scroll-a-thon for outdated details of projects past ranging from health games for kids to population models of shrimp.



Screen Shot 2019-09-04 at 5.54.14 PM.png

I’m a ringleader of sorts for CommonHealth, a project bringing together Cornell Tech, UCSF, Sage Bionetworks, and Open mHealth bring Personal Health Record support analogous to Apple Health(TM) to Android(TM) phones. Android Phone users will soon be able to securely download and view their electronic health record data, then share it with trusted apps and partners. We’re starting at UCSF, but the objective is to roll out to every institution in the US already connected to Apple Health. Stay tuned!


Curiosity Health

Apple's ResearchKit(TM) and ResearchStack for Android have gone a long way to standardize and improve mobile health app development. But unless you have your own dev team or $100k to burn, it's very unlikely you'll be able to build your own app for clinical work or research. Curiosity Health is building and leveraging open source tools including ResearchKit(TM) and ResearchStack to get the professional developer out of the loop. Currently, we're using our tools to build custom mHealth apps for clinical researchers around NYC in just a few days--at a fraction of the typical cost.


PAM: The Photographic Affect Meter


PAM is a visual, single-item measure of affect that has been validated as a proxy for gold standard measures of affect including PANAS. Although PAM started out as my dissertain project, the response to the project has been tremendous. PAM is open source and available for free to anyone, and has been used in clinical trials, behavioral research, employee/ workplace assessment, market research, and some other stuff I'm sure I don't know about. To get started with PAM yourself, get Open mHealth compliant downloads and source code on our small data lab omh apps page or visit the meager but informative PAM resource page. Or, shoot me an email. One of these days I'll put together the proper website it deserves!

Pollak, JP, Adams, P, Gay, G. (2011). PAM: A Photographic Affect Meter for frequent, in situ measurement of affect. Proceedings of CHI 2011, 725-734. (pdf | ACM Digital Library)


The Small Data Lab at Cornell Tech


My research home is the small data lab at Cornell Tech. We help people collect the digital traces they're already leaving behind and put them to good use (rather than just generating ad revenue for someone else). I work on several health-focused projects:

Supporting patients with chronic pain

Our work in pain covers a range of conditions from RA to chronic lower back pain. With Limbr, funded by UnitedHealth Group, we are working to decrease the likelihood of opiate use and addiction through behavioral mechanisms. We’re currently piloting a suite of apps to explore how small data such as physical activity traces can fuel peer coaching toward a goal of habituating an exercise regimen necessary for long term success without medication or surgery. Limbr builds on an NSF Smart and Connected Health project where worked to identify ‘behavioral biomarkers’ for pain through passively and actively sensing activity with mobile phones.

Selter A, Tsangouri C, Ali S, Freed D, Vatchinsky A, Kizer J, Sahuguet A, Vojta D, Vad V, Pollak JP, Estrin D (2018). An mHealth App for Self-Management of Chronic Lower Back Pain (Limbr): Pilot Study. JMIR Mhealth Uhealth, 6(9): e179 (full article)

AVA: Adaptable Visual Assessment

Healthcare professionals use Activities of Daily Living (ADL) to characterize a patient’s functional status and to determine and evaluate the effectiveness of treatment plans. ADLs are traditionally measured using standardized text-based questionnaires and the primary form of personalization is in the form of question branching logic. YADL, emphasis on Your, improves on this paradigm by personalizing ADLs to each patient and making reporting quick, engaging, and easy.

Yang, L, Freed, D, Wu, A, Wu, J, Pollak, JP, Estrin, D (2016). Your Activities of Daily Living (YADL): An Image-based Survey Technique for Patients with Arthritis. PervasiveHealth. (pdf)




I founded Wellcoin along with Glenn Laffel on the premise that if we could meet people where they are and encourage them to make just a few more healthy choices each day, we could make a very positive difference in their lives. In the app, members earned Wellcoins for doing healthy things, and they earned more the more 'proof' they had. Those Wellcoins were good gift cards and free products and services from health conscious partners like Whole Foods, DICK's Sporting Goods, and more. More importantly, Wellcoin was a thriving, health-focused community of tens of thousands of normal Bostonians sharing their struggles and triumphs trying to live a healthier lifestyle.

Wellcoin won several design and product awards, was a beloved local app, and insane engagement and usage statistics. Sadly, we failed trying to scale the concept nationally and had to shut down. I still believe in this idea, and someone should do it right. I’ll help!


Even older health tech projects

Pushcart: grocery tracking and recommendations

Fully integrated into online grocery services Instacart, FreshDirect, and Peapod, Pushcart is a small data email service helping users define nutritional goals and providing direct feedback on how their grocery purchases support their personal health goals. Pushcart works seamlessly with your online grocery service to analyze your purchases and map them to your personal health goals. No receipt scanning, no apps to install. Just clear insights and tips on achieving your health goals delivered straight to your email inbox.

Baum, A, Carroll, M, Estrin, D, Gunasekara, L, Pollak, J.P. (2015). Pushcart: Supporting and Scaling Nutritionist-Client Relationships. Proceedings of  CSCW 2015: Workshop on Moving Beyond e-Health and the Quantified Self, March, 2015.

VERA: mobile phone-based health behavior tracking and community

How can mobile phones can be used to employ various forms of motivation-both social and individual-to encourage healthy behavior? Intrinsic motivators are innate motivational factors such as competition, cooperation, control, and recognition that have been leveraged to bring about behavior change in many circumstances. Social influence has been shown to play an important role in persuasion and the motivation of behavior change; countless studies, both involving technology and not, have shown that individuals grouped with peers have better results in alcohol and smoking cessation, losing weight, exercising, and even surviving cancer. Through good design, all of these motivational factors can be employed and studied in mobile phone applications. We have developed and pilot-tested a health behavior application called VERA in which users take photos to document health-related behaviors, then rate and reflect on their own behaviors as well as those of their peers. Preliminary data show that individuals using VERA exhibit generally healthier behavior than those who don't, and that individuals using VERA with their peers are healthier yet. Furthermore, VERA collects volumes of highly valuable data, explicitly documenting the day-to-day health-related behavior with associated stress and emotional state of participants. This represents great promise for the use of cell phones as a means of encouraging healthier behaviors related to weight loss and the prevention of obesity.

Baumer, E.P.S., Khovanskaya, V., Adams, P., Pollak, J.P., Voida, S., Gay, G. (2013). Designing for Engaging Experiences in Mobile Social Health Support Systems. IEEE Pervasive Computing, 12(3), 32-39. (IEEE Pervasive)

Baumer, E. P. S., Katz, S. J., Freeman, J. E., Adams, P. J., Gonzales, A. L., Pollak, J. P., Retelny, D., Niederdeppe, J., Olson, C., Gay, G (2012). Prescriptive Persuasion and Open-Ended Social Awareness: Expanding the Design Space of Mobile Health. Computer-Supported Collaborative Work 2012. (ACM Digital Library)

VERA won multiple BOOM awards! BOOM is Cornell's student design competition. Congratulations to undergrads Vera Khovanskaya, Andy Liang, Andrew Ehrlich, and Stuart Davis who are really responsible for the existence and excellence of VERA and took home awards from Yahoo! and Google.

Aurora: mobile social support

Numerous studies have expounded on the benefits of social support for health, demonstrating improvements in overall quality of life, reductions in anxiety, reductions in perceived pain, increased comfort, and even faster recovery times. Unfortunately, not everyone has access to extensive social support networks or socially supportive therapy groups, and even those who do frequently don't have access when they need it most, such as when sitting in a waiting room or lying in bed awake late at night. To that end, we have created Aurora, a mobile phone-based application that allows patients to quickly and easily share their current mood and emotions with one another. Users select photos that they feel represent their current emotional state, and can likewise view the current emotional state of their peers. Aurora supports a range of forms of communication designed to help each user balance their own privacy with their desire for social interaction, and encourages them to reach out to one another in times of need or maintain distance as appropriate. Users can play games with one another that test their perceptions of the images selected by others, adding depth to the level of engagement and interactivity. This rich interaction, centered on the sharing of emotion, should help a greater number people reap more of the benefits of quality social support. Further, the emotions each patient records in the system on a daily basis represent new and highly frequent data points that can be evaluated by care providers and researchers, and the monitoring of patient emotions on a daily basis could help stave off many potential problems before they arise fully.

Gay, G., Pollak, JP, Adams, P., Leonard, JP (2011). Pilot Study of Aurora, a Social, Mobile-Phone-Based Emotion Sharing and Recording System. Journal of Diabetes Science and Technology, 5(2), 325-332. (JDST)

Mindless Eating Challenge/ Time to Eat!

Mindless Eating Challenge is a mobile phone-based health game based on Dr. Brian Wansink's Mindless Eating Challenge. In the game, players are tasked with caring for a virtual pet or plant, similar to the popular Tamgotchi. Pet care requires the user to follow a variety of health and eating recommendations and verify their actions with photos taken with their phone's camera. For example, the recommendation "Eat a hot breakfast" would require the player to submit a photo of him/ herself eating a bowl of oatmeal. Photos and compliance are then judged either by judges or peers. Based on compliance to these recommendations, the pet or plant changes its appearance and gains features or accessories--a tree might grow taller or grow more leaves or fruit in response. Alternatively, leaves might fall off if the players performance is poor. A social portion of the game allows the user to see various depictions of their performance in comparison to the performance of others in their group, as well as of their group in comparison to other groups. The game is designed so that various features can be easily enabled and disabled so it can be used as a platform from which to conduct research into the mechanisms of mobile persuasion in the context of improving health and well-being. This work was originally funded by the Robert Wood Johnson Foundation Pioneer Program.

Read an interview with me on healthGAMERS. Here is the original press release on NPR or Kotaku (for you gamers out there).

Pollak, JP, Gay, G, Byrne, S, Wagner, E, Retelny, D, Humphreys, L (2010). It's Time to Eat! Using Mobile Games to Promote Healthy Eating. IEEE Pervasive Computing, 9(3), 21-27. (IEEE Pervasive)

Byrne, S., Gay, G., Pollak, J. P., Retelny, D., Gonzales, A. L., Lee, T. & Wansink, B. (2012). Caring for Mobile Phone-Based Virtual Pets can Influence Youth Eating Behaviors. Journal of Children and Media, 6(1), 83-99. (T and F)


Wildlife Conservation

For many years (though not nearly many enough), I was deeply involved in modeling research for conservation biologists. I loved the work and the scientists I worked with regularly, including Bob Lacy, Phil Miller, Jon Ballou, Sara Zeigler, Philip Nyhus, and Becky Raboy. Below are super-brief descriptions of some of the projects I've been involved in over the years. Everything described below can be downloaded from this tools page at the Chicago Zoological Society.

Metamodel Approach to Wildlife Conservation Research

There are countless elements that go into creating an accurate model of even the simplest wildlife system--demographics, movement, disease, disasters, human intervention, etc. Most models simply choose to ignore or crudely approximate all but the one element that is currently being examined. A few models have attempted to incorporate as many of the factors as possible into one all-encompassing model (a megamodel). These approaches typically either ignore important characteristics or become too unweildy for most researchers to use.

The metamodeling approach, however, proposes that instead many smaller models be employed, each developed and run by the appropriate domain experts to carry out a very specific task such as modeling disease or deforestation. Not only is it likely that the outcomes of these models will better approximate the true outcome, but doing so will require extensive collaboration between researchers in many fields as well as stake-holders and local experts familiar with the system being modeled. The value of this collaboration cannot be over-looked in conservation research as typically all parties have to be completely on board with any recommended course of action for their to be any traction.

See papers below for much, much more. Funding for this work has come from many sources, including NSF, the Conservation Specialist Breeding Group (CBSG), the Chicago Zoological Society and various private donors.

Lacy, R.C., Miller, P.S., Nyhus, P.J. Pollak, J.P., Raboy, B.E., Zeigler, S.. Metamodels for transdisciplinary analysis of wildlife population dynamics. PLOS ONE, 8(12), p e84211.

Nyhus, P. J., R. Lacy, F. R. Westley, P. Miller, H. Vredenburg, P. Paquet, and J.P. Pollak. (2007). Tackling biocomplexity with meta-models for species risk assessment. Ecology and Society, 12(1): 31.

PMx: Software for pedigree analysis and management

PMx is a package of demographic and genetic analysis tools to assist with the management of breeding programs for wildlife species. PMx was developed by Jonathan Ballou (Smithsonian Institution/National Zoological Park), Robert Lacy (Chicago Zoological Society), and JP Pollak (Cornell University). The concepts and design of PMx benefitted immensely from input by the AZA Small Population Management Advisory Group, the EAZA European Population Management Advisory Group, the International Species Information System, and colleagues in zoos and conservation organizations around the world.

Lacy, R.C., Miller, P.S., Nyhus, P.J. Pollak, J.P., Raboy, B.E., Zeigler, S.. Metamodels for transdisciplinary analysis of wildlife population dynamics. PLOS ONE, 8(12), p e84211.

Spatial Modeling and Analysis of Wildlife Populations

Spatial, an agent-based, spatially explicit model, allows researchers to develop complex movement and dispersal simulations based on existing movement data (e.g. radio collar tracking data) or on easy to create rules. Landscape data can be imported from GIS for more robust models. Funding for Spatial was provided in part by CBSG.

Generalized Epidemiological Modeling of Wildlife Populations

Outbreak is a generalized, agent-based epidemiological model designed to allow researchers to approximate the effects of numerous classes of disease on a population. Outbreak models are developed through simple parameterization and requires no modeling or programming abilities to use. Funding for the development of Outbreak was provided by CBSG.