TIMWORK @timwork2008

1/ Ask a chatbot about a drawing and you get a plausible answer. In construction, "plausible" is how accidents happen. What practitioners need isn't an answer — it's a reason to trust it. That's the problem TIMWORK is solving. 2/ It starts with an ontology. We define members, specs, spaces, documents, accountability — and the relationship system between them — as a construction-domain ontology, then structure entire document sets into a knowledge graph on that schema. The agent doesn't free-associate: the ontology bounds its reasoning space, leaving structurally little room for hallucination. 3/ On top of it, our GraphRAG agent builds trust from three layers of evidence: 👁 Visual — every member in the answer lights up at its real location on the drawing 🔗 Relational — the ontological chain from member → spec → governing document is drawn live 🎞 Sequential — the agent's path to the conclusion replays sheet by sheet, step by step. 4/ That changes the moment of use. In a review meeting: "what's the basis for this girder spec?" — the answer appears with source sheets already on screen. Hours of digging become one question; a new hire navigates like a veteran. Asking, tracing, surfacing cross-document inconsistencies — it works right now. 5/ Every decision in construction management stands on documents. Turning them into agents that answer accountably — that's the market TIMWORK Inc. is opening. #ConTech #GraphRAG #Ontology #KnowledgeGraph #AIAgents

There's one reason construction teams don't trust AI answers: they can't see the evidence. At TIMWORK we built it differently: a GraphRAG agent grounded in a construction-domain ontology. Ask about a drawing — and watch which members, through which relationships, across which sheets led to the conclusion, step by step. You don't read the answer. You witness the verification. Not a demo — a tool you use today. #ConstructionTech #GraphRAG #Ontology #AI #Agent — @TIMWORKInc

AI Agent for Construction. Reading drawings. Reviewing documents. Finding risks. Searching information instantly. This changes everything.

Construction data was never truly searchable. Teams had to open drawings one by one to find information. But AI is changing that. With Vector Search, Graph RAG, and AI Agents, construction documents are finally becoming searchable. I believe this will change the future of AEC

I've launched a B2B SaaS that compares two construction blueprint images. Looking for feedback to improve it. Check it out at timdiff.web.app. Thanks!