The guided import reads what you already have rather than making you retype it. This article covers the first half: getting your files in and confirming how their columns were understood. For what happens after that, see Review and finish an import. If what you already have is a spreadsheet, what you gain by moving off one is the case for doing this at all.
Nothing is created during any of this. Items appear only when you approve them at the end.
Who can import
PI and Researcher roles. Junior Researchers do not see the import option.
Where to start
- On the Dashboard, the "Set up your lab" card has a Start setup button.
- Or in the top bar, open Add Inventory and choose Import inventory.
- Or go to Tools in the sidebar and open Import Inventory.
What you can upload
| Type | Notes |
|---|---|
| CSV spreadsheets | Export from Excel or Google Sheets as CSV. |
| PDF documents | Vendor order confirmations, packing slips. |
| Photos | JPG, PNG, or HEIC. A shelf, a freezer box, a set of labels. |
Excel files are not supported. Save the sheet as CSV first.
Upload everything at once
Click Choose files and add the whole batch: the spreadsheet, the PDFs, the photos, together. They are read into a single review rather than one import per file.
This is worth doing deliberately. Labs usually have their inventory scattered across several sources, and importing them one at a time means reconciling the overlaps yourself afterwards. Bringing them in together lets the review handle it in one pass.
You can also add another file later and have it fold into the same session.
The sample CSV
If you would rather start from a template than adapt your own sheet, use the Download a sample CSV link on the upload screen.
It has these columns:
Name, Quantity, Unit, Vendor, Category, Lot number, Received date, Expiration date, Storage location.
The single row inside it is a placeholder to be deleted and replaced. It is not real data and it is not a suggestion of what any value should be.
You do not have to use the template. If your spreadsheet has its own column names, the matching stage handles that.
Stage 1: Match columns
Once your files are read, the first step shows How I read your file: which of your columns it thinks maps to which Labsistant field.
Your job here is to check it and correct anything wrong.
What you will see
- Required columns. The minimum needed for a row to become an item. If one is unmatched, the step says which.
- A confidence indicator on each match, showing how sure it is that a column matches that field.
- No matching field on any column it could not place.
- Skipped, won't import these on columns being ignored, which is fine for columns Labsistant has no use for.
- Storage location is marked as set at the locations step, because that is handled separately and more carefully. See Review and finish an import.
Change any mapping that is wrong. This is the cheapest place to fix a problem, because everything downstream inherits it.
Fixes applied for you
There is a Fixes we'll apply for you section listing the tidying it did automatically, mostly unit normalisation, for example reading "grams" as "g".
Read it rather than skipping it. It is telling you what it changed, and it is the one place to catch a normalisation you did not want.
Units it does not recognise
If your file uses units Labsistant does not know, it says so and gives you two choices per unit:
- Change to one of ours, mapping it to a known unit.
- Keep this value as typed, leaving your unit as written.
Unrecognised units never block the import. You can proceed either way.
Whole items or a measured amount
Each reagent is treated as either whole items, such as boxes and packs, or a measured amount, such as a volume or a weight. Where this is implied by the unit you picked, the step says so and does not ask again.
Asking the AI for a closer read
If your column names are unusual and the automatic matching struggled, you can ask the AI to take a closer look at the file.
If it cannot be reached, the step tells you plainly that it could not reach the AI and lets you try again. It does not silently fall back to a guess.
Either way, the AI's reading is a proposal, not a decision. Everything it suggests lands in the same review you were going to check anyway, and nothing is created until you approve.
What is never guessed
Whatever route your data takes in, the import will not invent a compliance value. Lot numbers, CAS numbers, hazard levels, and expiry dates come from your files or from you, never from an assumption.
Rows missing something required are flagged rather than filled in. See Lot numbers, CAS, hazard and expiration.