Announcing Regen AI
[Week 1/12] Regen AI Update: Foundation & Kickoff - November 18, 2025
- Posted by: Shawn Anderson (Gaia AI)
- Development Phase: Phase 2 (Nov 2025 - Jan 2026)
- Key Focus: Launching weekly updates and introducing our core MCP infrastructure for planetary intelligence
Welcome to Regen AI Weekly Updates! 
Hello Regen community! We’re excited to launch this weekly update series to share our progress, challenges, and vision for Regen AI. As Gregory emphasized in our recent discussions, the forum is our central knowledge layer - this is where we’ll build context, invite collaboration, and document this incredible journey together.
Over the next 12 weeks, you’ll get a front-row seat to the development of what we’re calling “planetary intelligence infrastructure” - AI systems designed to amplify ecological regeneration by making data legible, processes efficient, and collective intelligence accessible to everyone in the Regen ecosystem.
What is Regen AI?
Regen AI is the collaboration between Gaia AI and Regen Network, launched in August 2025 with a bold mission: to fuse artificial intelligence with natural intelligence, creating a “legibility layer” for environmental data, ecological credits, and regenerative action.
Our Vision
We’re building toward the Symbiocene - a post-Anthropocene era where humans, technology, and nature coexist in symbiosis. Rather than AI for AI’s sake, we’re creating AI for Earth’s sake, with every tool optimized for Planetary Return on Investment (PROI) - maximum ecological and social impact per dollar spent.
Our Partnership
This alliance unites:
- Gaia AI’s cutting-edge agentic AI technology and community infrastructure
- Regen Network’s established ecosystem for high-integrity ecological credit origination
- Shared values of open collaboration, commons-based stewardship, and the $REGEN token as our unified coordination asset
This is the formation of a deep ecosystem alliance grounded in shared vision and collective intelligence.
The Foundation: Three Core MCP Servers
We’re building on the Model Context Protocol (MCP) framework to create three specialized servers that form the backbone of Regen AI. Think of MCP servers as tooling for AI agents - they provide resources (data access), tools (function calls), and prompts (predefined workflows) that agents can intuitively understand and use.
1. KOI MCP - Knowledge Organization Infrastructure
What it does:
- Aggregates 15,000+ documents across the Regen ecosystem
- Combines semantic search (vector embeddings) with graph queries (RDF triples)
- Pulls knowledge from Discourse forums, GitHub, Medium, Notion, websites
- Generates daily and weekly digests of network activity
How it works:
Active sensors function as data scrapers, continuously collecting information from all these sources. This data flows through the KOI network, gets vector-embedded using BGE embeddings, and populates a PostgreSQL database for semantic search. Simultaneously, the knowledge transforms into graph data stored in an Apache Jena server with 3,900+ RDF triples.
Why it matters:
Any Regen AI agent can now search the entire knowledge base semantically (“what projects increased soil carbon in tropical regions?”) or traverse the knowledge graph to understand relationships between concepts, methodologies, and projects.
Current capability:
- Daily digest analysis of network updates
- Weekly podcast generation
- Comprehensive searchable archive of Regen knowledge
2. Regen Ledger MCP
What it does:
- Queries on-chain data from Regen Ledger
- Resolves IRIs (Internationalized Resource Identifiers) from the data module
- Provides agents with real-time information about eco-credits, methodologies, projects
- Enables AI to understand the state of the regenerative economy
How it works:
Built on excellent groundwork from Jeancarlo, with planned expansion (mapped with Marie) to fully resolve IRIs from the data module. Agents can ask questions like “how many carbon credits have been issued in the past month?” or “what’s the current supply of biodiversity credits?” and get verified on-chain answers.
Why it matters:
This connects AI intelligence directly to the source of truth - the blockchain. Every eco-credit transaction, every project registration, every methodology update, every governance proposal becomes queryable by intelligent agents. This is the foundation for AI-enhanced governance, automated reporting, and data-driven decision making.
Integration potential:
With the upcoming IBC 2 upgrade (Inter-Blockchain Communication Protocol 2), we’ll have trustless, permissionless bridging to Ethereum. This means Regen Ledger accounts can be called and operated by Ethereum addresses, and our AI agents can interface with the broader DeFi and Web3 ecosystem.
3. Registry Review MCP
What it does:
- Automates document verification for new project onboarding
- Provides a 7-stage workflow from initialization to completion
- Assists registry reviewers with completeness checks, evidence extraction, and cross-validation
- Targets 70% reduction in review time
How it works:
The workflow stages are:
- Initialize - Create session, load checklist template
- Document Discovery - Scan and classify all project files
- Evidence Extraction - Map requirements to document evidence
- Cross-Validation - Check consistency across documents
- Report Generation - Populate checklist with findings
- Human Review - Present flagged items for expert assessment
- Complete - Finalize and export final report
Why it matters:
This is where AI meets real-world impact today. Becca and the registry team currently spend hours manually copying data between documents, checking for completeness, and cross-referencing requirements. This MCP does the heavy lifting, freeing humans to focus on high-judgment decisions and edge cases.
Development status:
We’re in Phase 2 with intense focus on the Registry MCP through January 2026. Early wins include automated document classification and metadata extraction. Next up: evidence snippet extraction with page-level citations.
The Architecture: How It All Fits Together
Any agent can use multiple MCPs. For example, the Registry Review Agent uses the Registry MCP for its workflow, plus the KOI MCP to search for methodology documentation, plus the Ledger MCP to verify project IDs.
The Double Quantum Leap
As Gregory mentioned in our November community call, we’re experiencing a “double quantum leap” - two orders of magnitude increase in network functionality:
- Ledger Upgrade (v0.53) - Enables IBC 2, emissions to different wallets, Ethereum interoperability
- New Roles Software - Multi-stakeholder organizations, role-based permissions via DaoDao
- Regen AI Integration - All three MCP servers + specialized agents
The convergence of these three initiatives enables:
- An eco-credit project creates a DAO through the new roles system
- The Registry Agent processes their documentation via Registry MCP
- Community members query the project’s status via agents using Ledger MCP
- All knowledge feeds back into KOI for future learning
Everything working together creates exponentially more value than any single piece.
Discussion Question
What aspect of Regen AI are you most curious about?
More specifically:
- Which of the three MCP servers interests you most and why?
- What use case would make the biggest difference for your work?
- Are you interested in beta testing? If so, which features?
- What questions should we answer in future updates?
- What workflows could AI help automate in your Regen work?
- What questions do you wish you could ask an AI about Regen Network?
- What data or insights would make your decision-making better?
- What pain points slow down your regenerative projects?
Let’s build planetary intelligence together! 
What to Expect in Future Posts
Every update in this 12-week series will include:
- Architecture and Strategy - Sharing our Vision for Regenerative AI
- Technical Progress - What we are building, what we are learning
- How to Participate - Specific ways to get involved
- Looking Ahead - Preview of next week’s focus
Looking Ahead: Week 2 Preview
Next week we’ll dive deep into the KOI MCP - the knowledge brain of Regen AI:
- How semantic search works with vector embeddings
- What our active sensors are collecting and from where
- Live demo of graph queries and knowledge traversal
- First auto-generated weekly digest of Regen network activity
- Plans for podcast automation with Amanda and Christian
Get ready to see how 15,000+ documents become intelligently searchable and actionable!
This is the first of 12 weekly updates documenting the development of Regen AI. Subscribe to this thread or the Regen AI section to get notified of new posts. All updates will be indexed in the pinned “Weekly Updates Index” thread for easy reference.