Makula - AI Copilot
An in-context assistant that gives technicians a sourced answer about a specific asset, pulled from that asset's own history — instead of a search bar and a stack of documents.
Overview
AI Copilot is an in-context assistant inside Makula that gives technicians a sourced answer about a specific asset, built from that asset’s own history — instead of a search bar and a stack of documents.
Problem
Technicians fix machines using knowledge that lives in their heads, not in the platform. When that person moves on, the fix goes with them — and the next technician starts over.
Solution
AI Copilot closes that gap. A technician describes the problem in plain language and gets back an answer built from that specific asset’s own history: past work orders, manuals, and recorded fixes.
Ask in plain language, get asked back
Copilot opens with a greeting and a few starting prompts instead of a blank field, since most technicians open it mid-repair without a database query in mind. A vague report — “I’m having an issue with my packaging asset” — gets a clarifying question in return before Copilot searches anything, so the eventual answer stays tied to the real symptom instead of a generic guess.
Steps to fix it, and where they came from
Once Copilot has a specific detail, it explains the fault and lays out numbered steps to resolve it, each carrying its own source marker — something to act on, not a paragraph to interpret. A Relevant Sources list sits just below with the manual, repair record, or call recording the diagnosis drew on, one tap away without leaving the chat.
The right asset, and nothing lost afterward
A switcher lets the technician point Copilot at one specific asset, with recent ones surfaced automatically, or fall back to Auto to search everything at once. Past conversations are saved by what was actually discussed, renameable or deletable, so a fix worked out today is easy to find again months later.