System will host its first public hackathon in New York on November 13-14, 2026, asking teams to turn scientific evidence into analyses, visualizations and working tools. The System Public Challenge will take place at the company’s headquarters, with a Friday reception and Saturday competition. Its raw material is the beta System Public Graph, a connected representation of findings from scientific research.
The assignment reaches into housing, employment, women’s health, malaria, public policy and city data. That range gives the event its interest: builders will have to carry evidence across disciplinary boundaries while keeping its meaning intact. A relationship that looks convincing on a screen still needs enough context for another person to judge whether it belongs in a decision.
For technical founders and research operators, the weekend offers a bounded opportunity to examine that translation. System is providing a common starting point; teams will supply the interfaces, analytical choices and explanations. The quality of those handoffs will determine how useful a prototype becomes after the pitch is over.
The System Public Graph represents statistical relationships extracted from open scientific sources, according to System. The company describes it as a beta offering containing millions of quantitative relationships. Its public materials distinguish the open graph from the broader System Graph, so developers should check which data and access arrangements apply to their project.
That structure changes the developer’s starting point. A team can ask how findings connect through shared variables, then investigate the studies behind those connections. It also creates design responsibilities: population, measurement, uncertainty and provenance need to travel with a result. An association between variables cannot automatically become a claim that changing one will cause the other to move.
System’s October 5 essay on making scientific knowledge computable sets out the company’s argument for persistent, connected evidence infrastructure. The business implication is practical. A tool becomes more useful when its next user can inspect how it reached an answer, understand the limits and continue the work without reconstructing the source trail.
There is relevant history behind that concern. System announced its integration of the Retraction Watch Database in 2023, an example of treating a finding’s status as part of the infrastructure. Builders at this challenge can carry the same discipline into their own interfaces: make the condition of the evidence visible alongside the conclusion.
The official challenge document is already available and gives teams concrete starting questions. One asks how indoor interventions connect to health and other outcomes; another examines relationships between job quality and wellbeing. These briefs create room for domain knowledge to influence what gets measured, compared and displayed.
The visualization assignments include mapping the distribution of women’s health evidence and comparing malaria prevention findings across settings. A useful submission could make uneven research coverage legible to a reader who lacks specialist training. That requires care with categories and denominators: a dense patch of findings and a large burden of need describe different things.
Build challenges extend into UK policy theories of change, discovery across World Bank indicators, alerts about malaria resistance evidence and comparisons between research and New York City data. These are assignments for investigation, rather than established results. Teams will need to distinguish a relationship supported by a study from a proposed connection that still needs checking.
Explore, Visualize and Build therefore give analysts, communicators and software engineers different ways to contribute. For a team choosing a project, a narrow question with an inspectable answer may be more productive than an impressive tour of every available relationship. A missing connection can itself guide the next research step, provided the interface explains whether the gap lies in the graph, the underlying literature or the team’s search.
The official event page lists Chethan Sarabu, Stefaan Verhulst, Daniel Gil-Sanchez, Travis Rich and Nalini Edwin as judges. Their roles span clinical innovation, public problem-solving, development data, knowledge infrastructure and communications. The lineup makes explanation part of the work: teams will need to show both what the prototype does and why its use of evidence deserves attention.
Cornell Tech identifies Chethan Sarabu as Director of Clinical Innovation for its Health Tech Hub. Stefaan Verhulst is co-founder and Chief Research and Development Officer of The GovLab at NYU. The World Bank’s profile of Daniel Gil-Sanchez places his data science and engineering work in its Development Data Group.
Travis Rich co-founded Knowledge Futures and returned as Executive Director in March 2026, according to the organization’s April announcement. Nalini Edwin is listed by the event as System’s Head of Communications. These are judging affiliations; they do not establish a separate speaker program or institutional sponsorship.
The overview assigns equal weight to meaningful graph use, usefulness, technical execution and communication. That combination should encourage teams to spend time on who can use the result and what that person needs to understand. A developer can implement a query correctly while still leaving the user with the wrong impression of what the evidence supports.
The official agenda and registration page puts the event at System HQ, 123 Lafayette Street, 4th Floor, New York, NY 10013. Friday’s welcome reception is scheduled for 6-9 PM. Saturday doors open at 9 AM, hacking runs from 10 AM to 3 PM, pitches follow, and winners are scheduled to be announced at 5 PM; participants should confirm timing with the organizer.
The event is free, teams comprise 2-7 people and individual registrants can request help finding collaborators. System is funding $7.5K in cash prizes. The participation rules allow advance research and planning but require implementation code to be written during the event; submissions need a prototype, demo video and public repository.
Teams retain their created intellectual property, while work containing or derived from the graph inherits its CC BY-NC license under the posted rules. Founders considering later commercial use should review those terms before choosing data dependencies. The license boundary belongs in the project plan alongside the technical architecture.
For readers following the wider local build calendar, DevCuration’s coverage of Ground Truth’s live-web agents hackathon and NASA Space Apps NYC’s open-data challenge offers related context. At System HQ, the work will turn on which scientific relationships teams choose to expose, which uncertainties they preserve and how clearly another person can pick up the resulting tool.
The Explore, Visualize and Build tracks give research analysts, designers, communicators and software engineers distinct ways to contribute. Teams can combine scientific domain knowledge with data work and interface design around one of the published challenge questions.
The posted rules allow advance research, planning, sketches and wireframes. Implementation code must be written during the event, and submissions require a working prototype, a 2-3-minute demo video and a public code repository.
Participants retain ownership of the intellectual property they create, while work containing or derived from the graph inherits its CC BY-NC license under the posted event rules. Teams should review the underlying data and tool terms before choosing dependencies for a later commercial product.
System describes the graph as a representation of statistical associations extracted from scientific sources. A useful tool should retain study context, provenance and uncertainty, because an association alone does not establish a causal effect.
The event overview assigns 25% each to meaningful use of the Public Graph, insight or usefulness, technical execution and communication. Teams should explain the practical question their prototype addresses and the evidence behind its output.