Open City Labs

Matthew Bishop, Founder & CEO is a member of two national committees developing standards for closed loop referral (360x) and social risk factors (GRAVITY Project). A sought-after speaker, Matt has spoken at the PCMH Congress, Youth Assembly at the United Nations, South by Southwest Interactive, Health 2.0 Conference, NYU, Columbia University and Cooper Union. Matt was the youngest person appointed to the NYS Health Care Reform Advisory Committee, which advised the Governor’s cabinet on the implementation of the Affordable Care Act. A thought leader on civic and health tech, Matt worked with NYC Councilman Ben Kallos to research legislation that would modernize access to government social services and testified in front of the NYC Council regarding the legislation. In 2010, Matt worked at Volunteers of America to oversee close to $100 million in government-funded social service programs and previously at the Manhattan Psychiatric Center and Kirby Forensic Psychiatric Centers.

As COVID-19 has ravished communities, the workers deemed "essential" too often earn poverty wages while they are disproportionately infected and killed by the virus. Open City Labs connects people to a one-stop-shop for government programs and community services. Our software includes a responsive website and is built to enable any software to natively refer to thousands of healthcare or social service providers, screen for eligibility for 300, and apply for 100 government benefits across 30 state agencies. This benefits data is populated by DataKit. DataKit accepts as inputs text, PDFs, application forms and compliance forms, and outputs statistically derived data models, “no-code” websites, documentation, and the code required to send and receive the application form data. It learns associated eligibility rules resulting in 12X fewer human hours than required to manually code those rules. The purpose of DataKit is to dramatically reduce the cost of cross sector systems integration. 

In the USA, eligible Americans don’t receive $80 billion in government benefits because they don’t sign up. Even in an affluent state like New York, if people know they are eligible, applying for multiple programs means filling out duplicate information on different often paper/PDF application forms. A survey of 453 New York programs/forms showed that 1% could be accessed via an API (e.g. Medicaid/Medicare billing), 7% online, and 47% with PDF form. In order to make government programs and community services accessible, program and human service data must be standardized in a machine readable format so it can be sent seamlessly between healthcare, social service providers and government agencies. We need tools to go from PDFs application forms and unstructured text to standardized human service APIs to enable Electronic Health Records systems or Case Management systems to natively screen for eligibility and apply for programs, or refer, using existing data to automate these processes. To do this at scale we need to develop (1) data standards for Human Services and (2) tools to map data between different exchange protocols and identify data models that shares the same semantic meaning (e.g. do Full Name and Legal Name mean the same thing?)

Our Datakit tool accepts as inputs text, PDFs, Application, Referral Forms and outputs (AI) statistically derived data models, web forms that reflect the PDF, code and documentation for sending and receiving that data. To add program eligibility data at scale it “learns” the eligibility rules requiring 12X fewer human hours, so that it can output eligibility questions and determinations. Our web app provides automation of referral and eligibility for 300 New York State programs and applications for 100 programs that accept PDFs making food, housing, healthcare, transit, anti-poverty and transit programs more accessible. We are partnering with Health Information Exchanges to enable their thousands of member organizations to be able to use their electronic medical record data to automate eligibility and program applications autofilling up to 95% of the fields of many forms, without having to integrate with each government agency. The NYS integrated eligibility screens for eligibility for 25 programs and applications are not auto filled across agencies. We have portals for both staff assisting their clients/patients (Navigator) and people (CARE Portal). We have formed a partnership with Amazon to integrate homeless data with medical records helping states identify healthcare trends and COVID-19 hotspots advancing policy and program development.

The company's mission is to make "government and community services, individually personalized, universally accessible and one click away." While our technology aggregates data that serve a wide variety of human needs and populations, our immediate focus is to improve the lives of homeless people because: (1) this is a population that at increased risk of getting infected by COVID-19, (2) dying from COVID-19 and (3) there is a lack of population health data to inform public health decision making. Existing data on the health of homeless patients rely on surveys, interviews, or patient clinical assessments. By matching health and housing records and providing population health analytics policymakers have accurate data to inform allocation of testing, housing and healthcare resources, while making programs more accessible. I have spoken to numerous homeless people about what accessing government programs is like as well as service providers, which struggle with the paperwork. Among the comments I have received:

“Every person who runs my program nationally needs your software!”

Katie Clay, Director of Medicaid Health Home Program, Hudson River Healthcare

“How much would it cost if the state bought it and made it free for every nonprofit?”

-Allison Sesso, Executive Director, Human Services Council

Our technology makes proven programs accessible to elevate opportunities for all people. In particular for communities that are left behind, lack of awareness, complex/burdensome applications and shame prevent people from accessing programs. Our focus on recommending programs based on people’s goals and barriers is consistent with the patient centered model of care management that is advanced in medicine. Our hope is that it will change beliefs, lessening the shame around accessing government programs by showing them that accessing government benefits is just a part of helping them achieve the goals that they have for themselves and their families.

In November 2016 I wrote a blog post that became the central idea of my application for the TED Residency Program. When I became a TED Residency Finalist, during the interview I was told that this was an “idea worth sharing,” but I would need to show some traction before giving a TED talk. One hackathon later I had designs and a meeting with Katie Clay, the Director of the Medicaid Health Home Program at Hudson River HealthCare, a care management program serving patients with multiple chronic conditions across an 8-county region in New York. “Every person who runs my program nationally needs your software,” she said after seeing the designs. Little did she know I hadn’t written a line of code. After getting the following reaction from Frank Kruppa, Commissioner, Tompkins County Health Department, “This is clearly the future! If the other departments embrace this software, I will be behind it,” I moved into my parents basement to build the prototype. I reconnected with my childhood friend Noah Whitman who had finished working as a Data Scientist for Cleveland Clinic. Brian Handspicker took over technical architecture work later when we met while working on the Gravity Project.

“Tell my story. I want my grandchild to know that I exist,” said Felicia Good. Felicia had told me her story of being a victim of child abuse, assault, rape and, not surprisingly, substance use after I helped her move her cart, which had gotten stuck on the trolly tracks in San Diego, California. Felicia had become homeless when she lost her Section 8 housing subsidy after her son moved out of the house. Unfortunately, because housing records are isolated from healthcare records, even the top public health experts do not know how many homeless people have diabetes, or COVID-19. While surveys show higher rates of hospitalization, mental health and drug us, our aggregate population health data is bad because it relies on clinicians to ask patients about their housing status and report it. For over 500,000 people in the United States, who like Felicia experience homelessness on a given night, policy/program development must begin with a basic picture of the health, housing, work, education and other factors shaping their well being.

Our team has the unique combination of healthcare, government and social services domain expertise, as well as information architecture, AI and Natural Language Understanding expertise to deliver on this project. 

Brian Handspicker, Chief Technology Officer currently a technical leader on the National Interoperability Collaborative Project Unify initiative – an effort to enable interoperability between health, human services, education and justice domains in support of improved individual well-being and population health. He is a subject matter expert on mapping and modeling of information between clinical healthcare standards and non-clinical National Information Exchange Model (NIEM) Domains such as Justice and Children, Youth and Family Services. He is also an expert on privacy and security of sensitive and personally identifying information (PII). He serves on the NIEM Technical Architecture Council (NTAC) and as a representative for NIEM Health on the NIEM Business Architecture Council (NBAC). 

Noah Whitman, Senior Data Scientist has extensive experience in data science, natural language processing and understanding and machine learning at some of the world’s most prestigious hospitals and research institutes. This includes leading studies that leverage unsupervised learning cluster analysis techniques at Cleveland Clinic and research in bioinformatics, plant ontologies. At Cleveland Clinic he led an analysis to understand whether high-cost patients were clustered in a few group of “Hot spotters,” or diffused throughout departments. This research led to an article I coauthored, “High-Cost Patients: Hot-Spotters Don’t Explain the Half of It,” which was published in the Journal of General Internal Medicine.

On January 6, 2020, Open City Labs submitted a Small Business Innovation Research grant application to pilot our technology to transform the NYC Department of Homeless Services (DHS) PDF referral form into a web form that can be autofilled by integrating with Healthix, an arm of the state health information exchange. The platform would facilitate care transitions, referral and program applications for homeless patients that are being discharged from healthcare facilities into homeless shelters. To secure a non-binding letter of support for the grant application staff had to get multiple agencies to sign off. As we scheduled a final presentation to secure a minimal agency funding in case the grant was not approved, Washington saw its first cases of COVID-19. Our meeting with DHS was cancelled as NYC saw a rise in cases. Knowing that projections showed that the peak of cases were expected to arrive in early April, we submitted an offer to stand up a minimum viable product in 3 weeks so that COVID-19 homeless could be safely transitioned to shelters where they could socially distance. There was no response. 

The Gravity Project brings industry leaders together to identify and harmonize social risks and protective factor data for interoperable electronic health information exchange. In March 2020 as COVID-19 cases grew rapidly in the USA I sounded the alarm about its potential impact on the homeless population. In response, the Gravity Project set up a COVID-19 response page and I joined the Homeless Management Information System-Health Information Technology Workgroup which I urged to leverage existing housing data standards to exchange housing data across health information exchanges. As a result of the committee’s work a shelter in Buffalo is now receiving alerts when their clients are diagnosed with COVID-19. In June we partnered with Amazon, which recently won a contract to aggregate all of  New York State’s electronic medical records over 15 million across the state of New York. Amazon agreed to help us set up a meeting for us to jointly pitch the NYS Department of Health. Our pitch is focused on: (1) a way of informing clinicians that their patient is homeless, our referral tool and our Homeless Population Health tool, which enables public health leaders to look at patient diagnosis or symptoms by patient housing history, or program enrollment.

The OCL Community Services Information Exchange and DataKit tools will bring together: (1) our existing AI trained models, (2) domain specific ontologies which define structural representation of the relationships between concepts and (3) open source knowledge graphs like Wikipedia’s DBPedia. While knowledge graphs take similar approaches to linking data (e.g. Resource Description Framework) ontologies like International Classification of Diseases (ICD-11) represent data from specific domains, essential for streamlining systems integration. By first connecting industry specific data standards across government, health and human services domains into one ontology, generating code and documentation dramatically reduces the cost of systems integration. Customers wishing to integrate will be able to move seamlessly between cross domain data models and protocols (HL7v2.5.1, NIEM, CDA, FHIR). Inspired by advances in the field of bioinformatics (Ma N., Zheng HT., Xiao X. 2017) , we intend to apply text mining to knowledge graphs like DBPedia and government ontologies like the Metathesaurus, a large biomedical thesaurus developed by the National Library of Medicine that links concepts and their relationships across more than 200 source vocabularies. We expect this will enhance the robustness and accuracy of our APIs. We will also compare performance of our algorithms against the leading open source candidates that are found on the General Language Understanding Evaluation (GLUE) Leaderboard. We will train AI algorithms using a more generalized (non government/clinical) dataset--like the Quora Question Pairs utilized in the GLUE benchmarks--to see if they enhance or reduce DataKit’s performance.

Open City Labs (OCL) is working with our partner Stewards of Change Institute (SOCI) to effect real change within communities based on their “Human Services 2.0 Theory of Change” using the OCL Community Services Information Exchange. Over the past five years, Stewards of Change built, tested and fine-tuned their Theory of Change model with the generous input and feedback from hundreds of passionate leaders who have participated in conferences and other SOCI convenings. SOCI has aggregated, synthesized and refined the thinking to create Human Services 2.0, our enhanced Theory of Change.

Human Services 2.0™ is a conceptual architecture that portrays the “To Be” vision of an interoperable ecosys­tem that includes health, education, human services and other consumer-oriented social services. It describes the long-term image, or future state, of what a connected and coordinated human services system could or should look like:

›› It is a “theory of change” that can guide the development of interoperability initiatives across business, organization and technical levels to achieve better client outcomes and improved operational efficiencies.

›› It is customer-centric, family-focused, community-based and technology enabled.

›› An engaging and powerful collaborative experience for creating a graphically illustrated “Change Vision Landscape” and “Roadmap” to guide the overall interoperability process for leaders and stakeholders.

›› A common language and vocabulary, including 10 core “drivers” describing the range of business, organiza­tional and technical factors required for interoperability. We have captured the driver concepts in unique visual icons to make it easy to share them in a succinct, impactful way.

1. Creating a Consumer-Centric Focus

2. Bridging and Integrating Service Silos

3. Building Open and Inclusive Processes

4. Managing Confidentiality and Privacy

5. Measuring Data and Performance

6. Building Public and Political Will

7. Developing Innovative Funding Streams

8. Redesigning Workforce, Workflow and Training

9. Articulating Governance and Leadership

10. Creating an Interoperable Technology Framework

›› A set of analytical tools, maturity matrices and templates that can be used to create an organizational baseline and readiness assessment for change across the three domains of policy, structure and practice as well as these 10 drivers.

In January 2019 HUD’s Annual Point-in-Time Count determined that 567,715 people were experiencing homelessness. These people represent a cross-section of America - from every region of the country, family status, gender category, and racial/ethnic group. They represent the most dire circumstances of the more than 19 million people suffering housing insecurity (paying more than half their monthly income on housing) across the United States.

According to a United Nations global survey  in 2005, an estimated 100 million people were homeless worldwide, with as many as 1.6 billion people lacking adequate housing (Habitat, 2015).

 In one year we anticipate OCL Homeless Transitions Navigator will be live in New York and California. New York (92 thousand homeless) and California (151 thousand homeless) have the highest numbers of homeless residents of all the states and territories. OCL Navigator plus the OCL Community Services Information Exchange have the potential of impacting the lives of over 15 million people with severe housing cost burden in NY and CA. In 5 years we expect to be in every state in the country impacting over 50 million people and planning international expansion in Latin America, Asia and Europe. 

We envision our impact not just through the lens of gaining product adoption, but also through our contribution to community driven national and international data standards. Within the next year the Open City Labs’ leadership will continue to play leadership roles in establishing human services in the healthcare domain through committees like the Gravity Project, as well as the National Information Exchange Model. Over the next 5 years, as Open City Labs expands globally, DataKit will be developed to enable the creation of CSIEs in new languages. The Stewards of Change Institute will work to create a governance model of domain specific Human Services/Social Determinants experts who are responsible for updating the ontologies with close coordination between HL7, WHO and other standards making organizations. Our mutual goal will be to evolve ontologies in a way that dramatically reduces the cost of integrating systems, advance research leading to an integrated model of human well being. This model of human well being would be updated for the spillover effects of how investments in one domain are predicted to impact another. For example, investments in Supportive Housing have been shown to reduce healthcare costs among homeless people while improving health outcomes. People could interact with this model of human well being by explicitly stating their priorities where governments should be spending their money.  which could be aggregated across a community and would output programmatic investments in their locality, state or country that would be needed to realize that person/people’s values.

Open City Labs’ partnerships make our company highly scalable. We have strategic relationships with the following organizations: 

Product Technical Assistance, Resell Sales Channel/Subcontract

• Amazon – AWS Partner Network

• CMA – NYS Medicaid Data Warehouse

• Google Cloud Advantage

Integration & App Store/Marketplace

• Epic - Electronic Medical Records

• Allscripts - Electronic Medical Records

• Health Information Exchange(s) 

• MaxMD – CDA, DIRECT messaging

Open Source Technology Partners

• HSLynk – Homeless Management Information System

Change Management & Data Governance & Community Engagement

• Stewards of Change Institute

Research has shown that 80% of health outcomes are shaped by social determinants of health: access to healthy food, adequate housing, poverty, health insurance. Open City Labs technology bolsters cross-sector partnerships between government agencies, healthcare and social service providers that are collaborating to address these cross-sector issues. Staff users have a variety of job titles (Case Manager, Care Coordinator, Shelter Services Coordinator to name a few) and their common role is to sign people up for programs, or deliver a service. They benefit from:

1. Recommendations of government benefits based on goals, barriers/needs, and eligibility.

2. Integrated eligibility screening for 300 government programs across over NYS & NYC 30 agencies.

3. Single point of signup for up to 100 government programs and autofill up to 95% of the fields in some application forms.

4. Spending ½ the time tracking referrals.

5. Increased participation in proven clinical programs that increase reimbursement directly (e.g. Medicaid Health Homes).

6. Improved patient health through social determinants-oriented benefits (e.g. access to healthy food, etc.), which leads to higher reimbursements as part of value-based care contracts (insurance reimbursements tied to health outcomes).

7. Keeping patients healthy and out of the emergency room where Medicaid and uninsured patients do not fully reimburse hospitals that are obliged to provide care.

8. Tracking service usage and outcomes of referrals (e.g. missed appointment), and health outcomes of patients across coalitions of healthcare and social service providers, providing data for program development and policy feedback.

We are currently working through our partnerships with Amazon and CMA Consulting, which have existing contracts with states like New York. These partners are providing technical and sales assistance, setting up meetings with New York State Department of Health and New York City government agency leadership. Our pricing follows a software-as-a-service model for healthcare, social service providers and government agencies. It is free for individuals.

Open City Labs has raised more than $300K and less than $1 million in convertible debt.

Open City Labs is raising $1-3 million in grant or equity investments over the next year. The amount we are raising will vary depending on where the company is with securing contracts.