1 of 7

MVP Architecture Optimization and Beta Product Scope

July 2025

@zoemcfox

@clarys_ai

2 of 7

What is Clarys.AI ?�Empowering OpenGov Decision-Making

  • AI assistant for OpenGov, enhancing decentralized governance and decision making.
  • MVP developed & validated with W3F Decentralized Futures grant.
  • Proven concept; seeking new funding to evolve into a full-featured, user-centric platform.
  • Analyzes on-chain & off-chain data (e.g., bounty submissions, forum discussions, AAG insights).
  • Improves governance quality, transparency & accountability.
  • Answers key stakeholder questions, such as:
    • Is the proposal credible and accountable?
    • Are there similar or related proposals?
    • What exists within a specific category?
    • Are wallet-level connections present?
  • Reduces complexity of sourcing and structuring data for impactful decision-making.

3 of 7

Problem Statement�Why Clarys.AI ?

  • OpenGov is complex, fragmented and time-consuming:
    • Manual effort needed to gather and interpret data from multiple sources
    • Limited visibility/accessibility for governance users (double dipping, overspending, proposal sourcing)

​

  • Clarys.AI simplifies governance by:
    • Automating data collection (on-chain & off-chain)
    • Streamlining analysis for faster, insight-driven decisions
    • Reducing manual workload and improving efficiency

​

  • Goal: Become the trusted source of truth for OpenGov data
    • Empower users to make informed, high-impact decisions quickly and confidently

​

4 of 7

Clarys.AI MVP Feedback�& Optimization Focus

Key MVP Challenges Identified:

​

  • AI hallucinations & lack of reasoning
  • Poor data retrieval; mixing unrelated info
  • Incomplete/false structured outputs
  • Missing links between proposals/entities
  • Obscure categorization due to inconsistent data
  • Weak handling of relationships & similarity queries

​

Planned Beta Optimizations:

​

  • Improve data normalization & structured storage
  • Combine vector search with structured retrieval
  • Use multi-agent, graph-based architectures
  • Add reasoning & self-reflection to reduce errors
  • Enhanced User Frontend

​

Goal: Deliver transparent, accurate and accountable AI support for OpenGov decisions

Private MVP testing with 20–25 OpenGov super users, received valuable hands-on feedback collected for Beta development.

5 of 7

Key Benefits of Using Clarys.AI

  • Instant Access to Critical Proposal Data:
    • Discover and understand on-chain proposals and bounty submissions in minutes, no more manual digging or fragmented sources.

​

  • Trusted Intelligence, Purpose-Built for Polkadot:
    • Make decisions based on a reliable AI trained exclusively on Polkadot-specific data, your single source of truth for OpenGov.

​

  • Achieve Over 80% Time Savings:
    • Streamline your governance research workflow and reclaim hours, focus on strategy, not information gathering.

​

  • Structured, Actionable Governance Insights:
    • Receive OpenGov data pre-categorized and organized for clarity, speed and confident decision-making.

​

  • Transparency on Demand:
    • Instantly request structured data on both on-chain and off-chain activities of OpenGov participants, promoting accountability and user transparency.

​

6 of 7

Milestones �Guiding Principles: Iterative development, minimized vendor lock-in, transparent AI decision-making, scalable, modular architecture

7 of 7

Thank You!

Clarys.AI is set to become an indispensable tool for simplifying and enriching the OpenGov experience,

While laying the groundwork for a more transparent, efficient and collaborative governance future.