Inspiration
Every year, roughly 30,000 people leave California’s prisons. Studies show that individuals who complete self-help and rehabilitation programs re-offend at less than half the rate of those who don't. Yet, less than 1% of California’s corrections budget goes toward these life-changing programs.
When we spoke with community organizers and Transformative Programming Works (TPW), a coalition of 120 community groups, we found a glaring bottleneck: many programs are run by single, overworked individuals. Attendance lives on paper sign-up sheets that end up lost in drawers, making it nearly impossible to prove enrollment and completions to state lawmakers.
What it does
We were inspired to build Spera (Latin for "to hope") to eliminate administrative friction for organizers and equip advocates with rock-solid, lawmaker-ready evidence to unlock critical funding.
How we built it
We designed Spera around two distinct user experiences—the local organizer in the field and the TPW coalition leader:
- Effortless Attendance Tracking: Organizers can drag and drop existing spreadsheets (with automated column mapping, even for Spanish headers) or tap quick, low-stress estimate range buttons (like 10–19 or 20–34) when exact numbers aren't available.
- Automated Grant Support: Organizers save core organizational details (EIN, address, signers) once, and Spera auto-populates state grant forms. An integrated AI assistant scans uploaded files against California Department of Corrections and Rehabilitation (CDCR) requirements to draft program summaries.
- Coalition Analytics & AI Reporting: TPW leaders get a high-level network dashboard showing total completions, outcomes, and estimated tax savings. With one click, an AI summarizer turns complex network metrics into three plain-language talking points and generates a one-page report for meetings with Sacramento lawmakers.
Challenges we ran into
- Volunteers by community organizers: Community organizers are under immense stress on site. Designing an interface that accommodates incomplete data without penalizing the user required careful UX trade-offs. We created a transparent "flagged estimate" system so rough numbers remain useful without compromising data integrity for lawmakers.
- Context-aware AI document analysis: Ensuring our AI assistant accurately checked uploaded documents against strict CDCR compliance guidelines without generating hallucinated summaries required rigorous prompt engineering and strict data constraints.
- Balancing Granularity with Simplicity: Building a dual-sided architecture that serves both micro-level organizer needs (grant readiness checklists) and macro-level advocacy needs (coalition-wide tax savings) without cluttering either view.
Accomplishments that we're proud of
Taking an idea that solves a genuinely important problem (like rehabilitation funding in California) and making its story clear, human, and persuasive
What we learned
Reciprocity Drives Adoption: You can't just ask grass-roots organizations to enter data into a void. By giving them instant value back—like automated grant writing and document tracking—they are actively incentivized to keep data flowing.
Meet Users Where They Are: Enforcing strict, rigid input rules fails in fast-paced environments. Accepting rough range estimates and paper sign-up sheet approximations opens the door for real-world usability.
Data-Driven Advocacy Works: Complex data analysis can be a major barrier during fast-moving legislative meetings. Translating granular statistics into plain-language talking points bridges the gap between grassroots effort and state funding decisions.
What's next for Spera
Hopefully, we can bring Spera beyond California, and impact the lives of one person at a time.
Log in or sign up for Devpost to join the conversation.