Participation Architecture - Final Grant Report

DAO Programs & Initiatives Domain Allocator Offerings (prev Questbook)
1

Participation Architecture - Final Grant Report

1. Executive Summary

Project: Participation Architecture - Governance Data Pipeline & Deterministic Triage Rules

Links:

Summary:
Participation Architecture is a developer-first REST API that normalizes Arbitrum DAO governance data and applies deterministic triage rules, so governance tools can prioritize what matters without building custom infrastructure. The system ingests proposals from Snapshot.org, applies a versioned rulebook (21 rules), and computes a Delegate Fatigue Index (DFI) - a 5-component deterministic score measuring governance workload burden.

Both milestones have been completed and all deliverables shipped, including a live demo endpoint, full documentation, 3 video tutorials, and a tagged release.

Impact of the grant:
This grant enabled the project to move from a research prototype to a production-grade, documented API with a live demo. Without the funding, the rulebook formalization, fatigue index implementation, video tutorials, and production deployment would not have been feasible as a solo developer effort within this timeframe. The grant also funded the infrastructure (dedicated VPS) that keeps the demo API publicly available for the ecosystem.

2. Performance Against KPIs

Milestone 1 - Pipeline Hardening + Rulebook v1 + API v1 ($3,500)

KPI Target Actual Status
Reproducible Docker setup docker compose up works Docker + manual setup both work Met
Deterministic outputs with rule IDs API returns rule IDs + reasons Every response includes reasons array with rule IDs Met
Automated rule test coverage >= 20 rule cases 30 test cases (150% of target) Exceeded
Quickstart time-to-first-call <= 10 minutes ~5 minutes (clone, venv, install, ingest, run) Exceeded

Milestone 2 - Fatigue Index + Docs/Tutorials + Public Release ($3,000)

KPI Target Actual Status
Fatigue index reproducible Documented formula + weight config 5-component formula, fatigue_config.yaml, full documentation Met
Test suite Comprehensive coverage 55/55 tests passing (25 fatigue + 30 rule engine) Met
Video tutorials published 2-3 tutorials 3 tutorials on YouTube Met
Tagged release e.g. v0.1 v0.1.0 Met
Repo open-source and runnable Docker + docs MIT license, full docs, runnable by third parties Met
Live demo endpoint Best-effort demo pa.wyszomirski.online with SSL Met
Documentation UX >= 70% complete Quickstart in <= 30 min Quickstart tested at ~5 min; structured feedback collection ongoing Met
API p95 response time < 400ms for cached queries ~1.8s uncached (399 proposals processed per request). Caching planned for next iteration Partially met

Note on p95: The current implementation processes all 399 proposals through the rule engine on each request. Adding a response cache is straightforward and planned. The proposal noted “cached feed queries” - the architecture supports this but caching was deprioritized in favor of shipping all other deliverables.

Deliverables Summary

Deliverable Status
FastAPI REST API with stable schema + OpenAPI Shipped
GET /proposals/feed - normalized proposals with scores/labels/reasons Shipped
GET /proposals/{id} - single item with rule audit trail Shipped
GET /delegates/{address}/fatigue - fatigue index + components Shipped
GET /delegates/{address}/fatigue/history - audit trail Shipped
GET /health - service health Shipped
Deterministic Rule Engine + Rulebook v2.7.0 (21 rules) Shipped
Delegate Fatigue Index (5-component deterministic formula) Shipped
FatigueSnapshot persistence to DB Shipped
OpenAPI/Swagger interactive docs Shipped
Quickstart guide Shipped
API Reference documentation Shipped
DFI deep dive documentation Shipped
Python integration example Shipped
TypeScript integration example Shipped
Video Tutorial 1: Quickstart Published
Video Tutorial 2: Notification Bot Published
Video Tutorial 3: Customize Rulebook Published
Runnable demo scripts (scripts/) Shipped
Tagged release v0.1.0 Published
Live demo API with SSL Live

3. Qualitative Impact & Community Feedback

Key non-quantitative outcomes:

  • The rulebook approach (versioned YAML + deterministic rules + test suite) demonstrates that governance triage doesn’t need AI/ML - transparent, auditable rules can do the job. This is a reusable pattern for any DAO.
  • The Delegate Fatigue Index provides the first open, reproducible formula for measuring governance workload. Every computation is stored to the database for audit.
  • All 21 rules return explicit reasons (rule IDs that fired) - no black boxes. This aligns with Arbitrum’s transparency values.
  • The project is grounded in PhD research (Design Science Research, WSB University) on DAO governance as sociotechnical systems, connecting practical tooling to academic rigor.

Alignment with Arbitrum SOS:

  • KR 7.3 (research on increasing participation): The DFI provides measurable workload signals that can be used to test what improves participation.
  • KR 7.4 (increase voting participation): The triage API reduces cognitive overload by prioritizing proposals.
  • Objective 6 (DAO efficiency): Standardized triage outputs reduce duplicated engineering across tools.
  • Objective 3 (home of builders): Open-source middleware with docs and tutorials lowers the barrier for builders.

4. Financial Summary

Category Budgeted Actual % of Grant
Engineering (pipeline + rule engine + fatigue index) $4,900 $4,900 75%
Documentation + Developer Enablement (docs, tutorials, examples) $1,200 $1,200 18.5%
Infrastructure & Ops (VPS hosting, domain, SSL) $400 $400 6.5%
Total $6,500 $6,500 100%

No meaningful difference from the original budget plan. Infrastructure costs are ongoing (~$5/month for the dedicated VPS hosting the demo API).

5. Future Plans & Continued Ecosystem Alignment

Maintenance commitment:

  • The demo API at pa.wyszomirski.online will remain live for at least 12 months post-grant.
  • The repository will remain open-source (MIT license) and maintained.

Planned improvements:

  • Response caching for feed queries (to meet the <400ms p95 target)
  • Periodic data re-ingestion from Snapshot.org to keep proposals current
  • Additional governance data sources (Tally) as connectors become available

Continued Arbitrum engagement:

  • The project serves as the technical artifact for my PhD dissertation on DAO governance (WSB University), ensuring continued development.
  • A second grant application (Governance Resilience Toolkit, $14,000, Education domain) has been submitted, building on this work with video education content and an IRL workshop at WSB University.
  • Open to collaboration with other Arbitrum governance tools that want to integrate the triage API.

Public good commitment:
This tool will remain open-source forever. No token, no paywall, no data monetization.

6. Additional Remarks

  • The entire project was built as a solo developer effort, demonstrating that meaningful governance tooling doesn’t require a large team or budget.
  • The deterministic approach (no AI/ML, explicit rules, full test coverage) was an intentional design choice - governance infrastructure should be auditable and predictable.
  • The video tutorials use AI-generated narration (ElevenLabs) and automated terminal recording (VHS by Charm), making them reproducible and easy to update as the API evolves.

Powered by Arbitrum. Arbitrum Everywhere.

2 Likes