AI maintenance assistant
Technicians ask about a fault code, a part or a procedure. Fixxit answers from your company's manuals, schematics, parts lists and service notes, and cites the exact page. One tap opens the PDF right there. What your technicians learn along the way is saved to the machine, so the next person starts from it.
Reads what you already have
Knowledge retention
No more loss of information when troubleshooting. Part changes, settings and fixes that worked are saved against the machine, so every fix makes the next one faster.
Mention in chat that a part was superseded, a setting changed or a machine is running limited, and Fixxit offers to save it to that machine's knowledge base. Each entry is marked as from the manual, reported by a technician, or verified by an admin.
When a technician taps "Fixed it", the conversation becomes a field insight. Managers review it, and Fixxit can bring it up the next time that fault comes up, next to the manual.
Every change keeps its earlier versions, who reported it and when. What a senior technician knows stays with the machine after a shift change, a new hire or a retirement.
How it works
No re-typing, no tagging marathons. Fixxit works with the documentation you already have, however it's organized today.
Drag in whole folders or ZIP archives: OEM manuals, schematics, parts lists, service bulletins and your own procedures. Fixxit reads every page, scanned ones included, and sorts each file by machine and document type.
From the office or at the machine on a phone. Describe the fault code, the part or the job. Focus on one machine, or ask across the whole plant.
Every answer cites the document and page it came from. Tap a citation to open the PDF at that page and check the diagram. Verified fixes go back in as field insights for the next shift.
Features
Fixxit doesn't answer from the open internet. It reads your library, and shows you where every answer came from.
Cited answers
Fixxit points to the exact page it used: the section of the manual, the schematic sheet, the line in the parts list. Tap a citation and the original PDF opens at that page, so a technician can check the diagram before picking up a wrench.
Bulk upload
Drop in folders or ZIP archives and their structure comes with them. Scanned pages go through OCR so they're searchable too. CAD files and other formats without text are kept as reference entries, so Fixxit can still point people to them.
AI auto-detection
Fixxit reads folder names, file names and content to work out which machine each document belongs to, what kind of document it is and what to call it. You see how confident it is, and you can change anything afterwards.
Document library
Every manual, parts list and service bulletin in one place, grouped by machine, with processing status at a glance. After an upload, Fixxit summarizes what it found: new machines, auto-detected files and anything that needs attention.
Machine focus
Pick a machine and Fixxit starts with its documents, while still drawing on shared material like lockout procedures and plant standards. Starter prompts cover the everyday jobs: diagnosing a fault code, finding a part number, walking through a PM, summarizing lockout steps.
Machine knowledge base
Each machine has a knowledge base next to its documents: part numbers, fault codes and settings from the manuals, plus the updates technicians report from the floor. Fixxit reads it before the manuals and says where each fact came from.
Beyond the manual
When a technician confirms what actually fixed a problem, Fixxit captures it as a field insight and can cite it the next time that fault comes up. Hard-won know-how stays when people move on.
Machine statistics, fault trends and resolution rates drawn from the questions your team asks, so you can see which equipment is eating the most time.
Runs in any modern phone browser with nothing to install. Large touch targets and short, step-by-step answers that are readable in a hurry.
Owners and admins manage machines and documents. Technicians read and ask. Everyone sees the same, current library.
Each person's conversations are visible only to them, so technicians can ask the questions they'd hesitate to ask out loud.
Replace a manual or reprocess a file and the next answer uses it. No retraining and no waiting for a vendor.
Who it's for
For technicians
Stop digging through binders, shared drives and old emails while the line is down.
For maintenance managers
Turn scattered manuals and tribal knowledge into something the whole team can rely on.
Security & data isolation
Fixxit is multi-tenant by design. Isolation is enforced at every layer, from the database to the AI assistant itself.
Industries
If your equipment came with a manual, Fixxit can help your team keep it running.
Presses, CNC machines, welding cells and conveyors, with OEM manuals, PLC programs and electrical schematics.
Tunnel washers, dryers, ironers and folders from several vendors, often with decades of documentation.
Fillers, ovens, mixers and CIP systems, where downtime and sanitation windows are tight.
Case packers, wrappers, labelers and palletizers across mixed-vendor lines.
Boilers, compressors, chillers and HVAC spread across buildings and sites.
Pumps, valves, dosing skids and instrumentation with long service histories.
FAQ
From the documents your organization uploads: OEM manuals, schematics, parts lists, service bulletins, procedures and your own notes. Fixxit's AI reads and reasons over that library, and every answer cites the document and page it used.
Fixxit is built to answer from your library and to show its sources, so it's easy to see what an answer rests on. If the citations don't back it up, your team knows to check with a senior tech or the OEM. As with any reference, verify safety-critical steps against the original document.
Practically anything. PDFs (including scanned pages, via OCR), images and multi-page TIFFs, Word, Excel and PowerPoint in current and legacy formats, OpenDocument, RTF, email (EML and MSG), HTML, CSV, JSON, XML, text and config files, CNC and PLC programs, and DXF drawings.
Files with no extractable text, such as DWG and other CAD formats, video or firmware, are kept as reference entries so they still show up in the library and Fixxit can point people to them. Upload whole folders or ZIP archives at once.
Fixxit runs on Anthropic's commercial API. Under Anthropic's commercial terms, content sent through the API is not used to train its models by default. Your documents are used to answer your own team's questions, and nobody else's.
No. Every machine, document and conversation belongs to exactly one organization, and each request to the assistant is confined to that organization's library. The Security page explains how that works.
Yes. Fixxit runs in the browser on phones, tablets and desktops, so there's nothing to install. Technicians can ask a question at the machine and open the cited page right there.
Most of it is gathering documents you already have. Upload the folders as they are: Fixxit extracts the text, detects the machine and document type, and your team can start asking questions as soon as processing finishes. You can correct any classification afterwards.
We scope pricing with each customer, based on your sites, team and document library. See Pricing or book a demo to talk it through.
Book a demo and bring the manual for a machine that gives your team trouble. You'll see how Fixxit answers from it, page by page.
