Intento Labs
Services & Capabilities · 15 slides
Intento Labs
Intento Labs
May 2026
Drawings, 3D models, documents, spreadsheets — the data your team produces every day, finally working for them. Through our products, or directly inside the AI tools you already use.
A specialist engineering team building production AI for the people who design, build, and operate the physical world.
Three products live and shipping. Real customers. Built and operated by us, end-to-end.
Small team. Every conversation is with the people actually building the system. Toronto-based, working with clients across North America.
Working with the National Research Council (NRC) on applied AI for the built environment.
P&IDs, ISOs, 3D models, vendor specs, contracts, line lists, takeoffs, takeoff sheets, change orders — every project leaks knowledge across formats and tools.
We read the formats your team actually uses — IFC, DWG, P&IDs, ISO drawings, RVT, vendor PDFs — not just text-flattened blobs.
Multi-agent systems that reason about your project the way an engineer would: spatial, structural, regulatory, contractual.
Stand-alone product, embedded inside your Claude / ChatGPT setup via MCP, or behind your own UI. Your call.
Diff specs across revisions. Cross-check vendor data sheets against project standards and industry codes — ISO 13703, ASME B31.3, API, CSA, your own internal practices.
Decades of project knowledge lives in spreadsheets. We turn that into a clean, queryable data layer — then put a natural-language interface on top of it.
The deliverable that consumes weeks of engineering time — generated in hours. Built from your P&IDs and 3D models, kept in sync as the design evolves.
A modern alternative to legacy ISOGEN-style tooling. AI agents read your 3D piping models and produce clean, annotated isometric drawings — ready for pipe-spool fabrication.
Quote a new project the way an experienced estimator would — but in minutes, not weeks. AI reads your 3D model and scope docs together, drives takeoff, and rolls it up against historical pricing from past packages.
We plug AI directly into the tools you already use. Keep your Navisworks model open, open a chat next to it, and ask in plain English — get takeoffs, summary tables, BOMs, clash explanations, fabrication packages straight from your federated model.
.nwd / .nwc with CADWorx, Plant 3D, Revit, IFC objects
Same agents, same data layer, two delivery modes — use one, the other, or both.
A purpose-built web UI we design and ship for your team. Project-aware, role-aware, tuned to the specific workflows your engineers actually do.
The same agents, exposed as MCP tools. Plug straight into the Claude your team already uses — no new app to learn.
Three production AI products, a paying government customer, and a working agentic data platform — already live.
Multi-agent AI assistant for IFC building models. Talks to your 3D model in plain English — query it, validate it, edit it.
Try it live →Our agentic knowledge engine — the data layer that makes engineering files queryable by AI. Powers everything else we build.
See the platform →National Research Council partnership delivering applied AI for the built environment. Independent validation and a real customer relationship.
More on partnerships →Buying pipe at Home Depot doesn't make a transmission pipeline. The pipe is the easy part — what moves gas across a country is the system around it: connectors, specs, monitoring, codes. Off-the-shelf AI is the same.
Foundation first. Then domain automation. Same Claude — expanding scope as your team and your data are ready.
Every engineer, PM, and estimator gets a chat-first AI for the messy 20–30% of their day that's nobody's specialty — drafting, summarizing, comparing, debugging Excel, explaining specs.
The chat box your team learned in Phase 1 stays the same — we connect it via MCP to your engineering reality. Same UX, expanding scope.
Intento Labs
Start with one use case. Prove the value in four weeks. Scale into a real system from there.