PANCAKE: AI-Native Geospatial Storage for Agriculture
An AgStack Project | Powered by The Linux Foundation
PANCAKE (Persistent-Agentic-Node + Contextual Accretive Knowledge Ensemble) is an open-source, AI-native data platform designed as Digital Public Infrastructure (DPI) for agriculture.
Vision: "An AI-native Operating System for Agriculture"
Quick Start
# Clone repository git clone https://github.com/agstack/pancake.git cd pancake # Make the script executable (only needed once) chmod +x implementation/setup_postgres_docker.sh # Set up dockerised PostgreSQL with pgvector bash implementation/setup_postgres_docker.sh or ./implementation/setup_postgres_docker.sh # Install dependencies pip install -r implementation/requirements_poc.txt # Run POC notebook jupyter notebook implementation/POC_Nov20_BITE_PANCAKE.ipynb
See: implementation/POC_README.md for detailed setup instructions
What is PANCAKE?
PANCAKE is "ChatGPT for spatio-temporal farm data" - query your agricultural data with natural language, not SQL.
Core Components
- BITE (Bidirectional Interchange Transport Envelope): Universal JSON format for agricultural data
- SIP (Sensor Index Pointer): IoT/Streaming layer for sensors and actuators
- MEAL (Multi-User Engagement Asynchronous Ledger): Collaboration persistence layer
- TAP (Third-party Agentic-Pipeline): Vendor integration framework
- PANCAKE Core: AI-native geospatial storage with multi-pronged RAG
Key Features
- ✅ Natural Language Queries: "What pests affected Field A last week?"
- ✅ Multi-Pronged RAG: Semantic + Spatial + Temporal similarity search
- ✅ Polyglot Data: One table stores all data types (observations, imagery, operations, etc.)
- ✅ Spatio-Temporal Indexing: Automatic spatial relationships via GeoID
- ✅ Immutable Audit Trails: Cryptographic hash chains for compliance
- ✅ Vendor-Agnostic: TAP adapters for any agricultural data source
Documentation Structure
Core Documentation (/docs)
BITE.md- Universal data format specificationPANCAKE.md- Core storage and AI architectureSIP.md- Sensor/actuator streaming protocolMEAL.md- Collaboration and audit ledgerTAP.md- Vendor integration frameworkSIRUP.md- Enriched data payload concept
Sprint Plans (/sprints)
SPRINT_1_USER_AUTHENTICATION_UPGRADE.md- OECD-compliant authentication (Weeks 1-12)SPRINT_2_ENTERPRISE_MIGRATION.md- Enterprise FMIS migration (Weeks 13-24)SPRINT_3_PAYMENTS.md- Digital payments integration (Weeks 25-36)SPRINT_4_DATA_WALLETS.md- Data wallets & chain of custody (Weeks 37-48)
Testing Profiles (/testing)
testing_EUDR.md- EUDR compliance testing scenariostesting_food_safety.md- Food safety traceability testing scenarios
Strategic Documents (/strategic)
PANCAKE_WHITEPAPER_DPI.md- Business white paperEU_HORIZON_GRANTS_STRATEGY.md- EU grant strategyEU_HORIZON_GRANTS_PROPOSAL.md- EU grant proposalOECD_AUTHENTICATION_ALIGNMENT.md- OECD compliance analysisopenagri_integration.md- OpenAgri integration guide
Implementation (/implementation)
POC_Nov20_BITE_PANCAKE.ipynb- Proof of Concept notebookPOC_README.md- POC setup instructionsrequirements_poc.txt- Python dependenciessetup_postgres.sh- PostgreSQL setup scriptmeal.py- MEAL implementationtap_adapter_base.py- TAP adapter base classtap_adapters.py- TAP adapter implementationsmigrations/- Database migration scripts
Podcast Series (/podcast)
- Complete podcast series for NotebookLM
- 11 modules covering all aspects of PANCAKE
Archive (/archive)
- Historical documentation and completed phase summaries
Roadmap
See: ROADMAP.md for complete 12-month sprint-based development plan
Current Status: Sprint 1 (User Authentication) - Planning
Timeline:
- Sprint 1: Weeks 1-12 (User Authentication)
- Sprint 2: Weeks 13-24 (Enterprise Migration)
- Sprint 3: Weeks 25-36 (Payments)
- Sprint 4: Weeks 37-48 (Data Wallets)
Key Use Cases
1. Natural Language Queries
# Query PANCAKE with natural language answer = pancake.ask( "What pests or diseases have been observed in the coffee fields in the last week?", geoid="field-coffee-123" )
2. Multi-Vendor Data Integration
# Fetch data from multiple vendors via TAP ndvi_data = tap_factory.fetch('terrapipe_ndvi', geoid='field-abc') soil_data = tap_factory.fetch('soilgrids', geoid='field-abc') weather_data = tap_factory.fetch('terrapipe_weather', geoid='field-abc')
3. Enterprise FMIS Migration
# Migrate FMIS data to PANCAKE migration_tool = MigrationTool(pancake_client, adapter) result = migration_tool.migrate_csv('fieldview_export.csv')
4. EUDR Compliance
# Generate EUDR compliance report eudr_report = eudr_compliance.generate_eudr_report(shipment_geoid='shipment-123')
Architecture
┌─────────────────────────────────────────────────────────┐
│ PANCAKE Application Layer │
│ (Natural Language Queries, Voice API, Mobile/Desktop) │
└────────────────────┬────────────────────────────────────┘
│
┌────────────────────▼────────────────────────────────────┐
│ PANCAKE Core (Storage + AI) │
│ • BITE Storage (PostgreSQL + pgvector) │
│ • SIP Storage (Time-series optimized) │
│ • MEAL Storage (Immutable ledger) │
│ • Multi-Pronged RAG (Semantic + Spatial + Temporal) │
└────────────────────┬────────────────────────────────────┘
│
┌────────────────────▼────────────────────────────────────┐
│ TAP (Vendor Integration) │
│ • Terrapipe (NDVI, Weather) │
│ • SoilGrids (Soil data) │
│ • Custom TAP Adapters │
└─────────────────────────────────────────────────────────┘
Technology Stack
- Database: PostgreSQL 14+ with pgvector extension
- AI/ML: OpenAI GPT-4 (embeddings and queries), local models supported
- Blockchain: Hyperledger Fabric (payments), Hyperledger Indy/Aries (identity)
- Languages: Python 3.9+
- License: Apache 2.0 (Code) | CC BY 4.0 (Documentation)
Contributing
PANCAKE is an AgStack project under The Linux Foundation. We welcome contributions!
Getting Started:
- Review
ROADMAP.mdfor current development priorities - Check
GOVERNANCE.mdfor contribution guidelines - Join AgStack community: https://agstack.org
Areas for Contribution:
- TAP adapter development
- Core PANCAKE improvements
- Documentation
- Testing
- Use case implementations
License
- Code: Apache 2.0
- Documentation: CC BY 4.0
Contact & Resources
- GitHub: https://github.com/agstack/pancake
- Documentation: See
/docs/directory - AgStack: https://agstack.org
- Email: pancake@agstack.org
Status
Current Version: POC (Proof of Concept)
Next Milestone: Sprint 1 completion (Week 12)
Production Target: After Sprint 4 (Week 48)
An AgStack Project | Powered by The Linux Foundation