One Database, Two Ways: REST APIs + Agentic AI with Gilhari & ORMCP

Stockx is an agentic supply chain intelligence platform that lets users operate the same six-table inventory database in two ways. Pathway A — Manual CRUD Routine inventory operations are handled through traditional data grids, with zero AI overhead. Pathway B — AI + Generative UI Users can ask questions in natural language. A LangGraph agent loop uses ORMCP and Gilhari to access governed business data and return actionable Generative UI, including live charts, KPI cards, Kanban boards, forms, and more instead of plain-text responses. This demo uses Microsoft SQL Server deployed on Azure. However, the Gilhari/ORMCP architecture is database- and deployment-agnostic: the same approach can be used with supported relational databases deployed in the cloud or on-premises. This video walks through the complete architecture and demonstrates how both pathways work together to create an intelligent and flexible supply chain management system. Project Repository: https://github.com/sudhanva-bh/Stockx Built On: Stockx is built on JDX, Gilhari, and ORMCP, technologies developed by Software Tree, LLC. ORMCP Documentation: https://github.com/SoftwareTree/ormcp... Gilhari: https://hub.docker.com/r/softwaretree... Software Tree: https://www.softwaretree.com Key words: Agentic AI, MCP, Model Context Protocol, ORMCP, Gilhari, AI Agents, Semantic Object Model, REST API, RESTful API, Database Integration, AI Database, Natural Language Database, SQL Server, Azure, Supply Chain Management, Supply Chain Analytics, Database Agnostic, Enterprise AI Acknowledgements: Special thanks to Software Tree, LLC for providing the JDX, Gilhari, and ORMCP technologies that formed the foundation of the zero-code data layer and MCP-based AI tool-calling architecture, as well as for their support throughout Practice School-I. #agenticai #mcp #modelcontextprotocol #aiagents #ORMCP #gilhari #supplychain #database