01 Scoping
Build an open-source kernel for state unemployment insurance
Explore whether states could share a common, open-source core for unemployment insurance rather than each procuring and rebuilding the same foundational capabilities independently.
Our hypothesis
A shared kernel could lower the cost of modernization, create a healthier implementation market, make policy changes easier to absorb, and give states a durable foundation they can adapt to their own needs.
What we are exploring
- The business case for building and maintaining a shared core
- What belongs in the common kernel versus state-specific implementation
- Governance, funding, and long-term stewardship
- How federal requirements, systems, and partners should shape the architecture from the start
- What would make adoption credible for an initial group of states
Where feedback is useful
We especially want to hear from state UI leaders, state CIOs and CTOs, federal partners, implementers, vendors, funders, and people who have tried shared or open-source government systems before.
Related priority: Build shared public digital infrastructure
02 Exploring
Create a startup accelerator for public-sector infrastructure
Explore whether the field needs a repeatable way to identify important gaps and launch new organizations, products, and public infrastructure to fill them.
Our hypothesis
Some critical capacity gaps may be easier to solve by creating new institutions than by asking existing organizations to stretch into roles they were not built to play.
What we are exploring
- Which problems are appropriate for incubation or acceleration
- What support a founding team would need from idea through launch
- How efforts would be selected, governed, financed, and eventually spun out
- How this model should complement existing funders, civic-tech organizations, and government partners
- Which candidate ventures would be useful tests of the model
Where feedback is useful
We want feedback from founders, funders, government leaders, field organizations, and people who have built or backed mission-driven ventures before.
Related priority: Turn implementation knowledge into field capacity
03 Exploring
Build hands-on AI capacity inside congressional committees
Explore whether congressional committees could pair embedded technical practitioners with serious model access and wraparound support to expand committee capacity and give staff firsthand experience of what AI makes possible.
Our hypothesis
Congressional committees need more technical capacity to understand, oversee, and legislate in an AI-shaped world. Pairing strong technical practitioners with committee staff could help with the work immediately in front of them while helping members and staff form their expectations of AI through direct experience with their own work rather than demos and secondhand accounts.
What we are exploring
- What kinds of committee work would benefit most from embedded technical capacity
- How to identify and select forward-deployed engineers who can succeed in a legislative environment, not just strong engineers in general
- What embedded engineers need to understand about congressional procedure, committee operations, policy development, oversight, ethics, security, and the norms of the Hill before they begin
- What ongoing technical assistance and program support committees and embedded engineers would need beyond the placement itself
- How model access and token or compute budgets should be provisioned, funded, and governed so teams can experiment meaningfully rather than merely demonstrate AI tools
- How to structure relationships with frontier model providers while preserving the independence, security, confidentiality, and public responsibilities of congressional committees
- Which committees would make credible first placements, and what a successful first cycle would look like
- How to measure both immediate value to committee work and longer-term gains in institutional technical capacity
Where feedback is useful
We especially want to hear from current and former congressional staff, committee leaders, technical-assistance organizations, engineers who have worked inside government, AI companies, congressional fellowship programs, ethics and security experts, and others who understand what it takes to introduce new technical capability into legislative institutions.
Related priority: Build permanent public talent and institutional capacity