CSV and file uploads + Orcabase

Upload an export. Ask it anything.

Got an export from your payment tool, your store or your courier? Upload it and it becomes something you can ask about, chart and compare with the rest of your data. No formulas, no pivot tables.

  • File upload
  • Nothing to set up
  • Files up to 200 MB
AgentCSV and file uploads connected

Which deliveries were late in August, according to the courier’s export?

Reading your CSV and file uploads data…

Monday deliveries were late most often: 21%, twice the other weekdays.

  • Nearly all were weekend orders waiting for pickup.
  • The regional courier carried most of the late ones.
  • Late orders drew three times as many support tickets.

Late deliveries by weekday, %

21MTWTFSS
How I got this· 4 steps
  1. Read the courier export you uploaded
  2. Compared promised and actual delivery dates
  3. Worked out the late share by weekday
  4. Matched late orders with your support tickets

Example answer. Numbers are illustrative.

In plain words

How can I analyze a CSV file with AI?

Upload it to Orcabase. Each column’s type is worked out from the data, so amounts add up and dates sort. Then ask about it in plain words and get the answer with a chart.

Unlike pasting a file into a chat, the upload stays next to the rest of your data, so the agent can compare it with your orders or payments, and your whole team sees the same numbers.

Ask in plain words

What you can ask about your CSV and file uploads data

The agent answers with the number, a chart and the steps it took. Keep the good answers on a dashboard.

  • What’s in this file, in a sentence?

  • What were total sales by month in this export?

  • Which customers are in this export but not in our store?

  • Which courier was late most often?

  • What changed between last month’s export and this one?

  • Which products are listed at different prices?

  • Show me the top 20 rows by amount.

  • Group this by region and chart it.

What comes in

Your CSV and file uploads data, ready to ask about.

A new table in your Orcabase warehouse, or in a database you connect.

01

The formats tools export

CSV and TSV from almost anything, plus Parquet and JSON, including one record per line.
02

Types worked out for you

In your Orcabase warehouse, numbers, dates, true/false and text are detected from the data.
03

Up to 200 MB a file

Room for years of orders from most tools.
04

Next to everything else

Uploads sit in your warehouse with your synced data, so one question can use both.

Numbers worth defining

Define them once. Every answer agrees.

Tell Orcabase what each number means, certify it, and your dashboards, the agent and your whole team use the same one. The agent can draft these from your CSV and file uploads data.

How metrics work

  • On-time delivery

    Share of deliveries that arrived by the promised date.

  • Cost per delivery

    Courier charges divided by deliveries, from the courier’s export.

  • Marketplace revenue

    Sales from a marketplace export, minus its fees and refunds.

  • Return rate

    Returned orders as a share of orders shipped.

Dashboards to start with

Your CSV and file uploads numbers at a glance.

Ask the agent to build one of these, or pin charts yourself. Share any dashboard with a link.

Deliveries

3 charts
  • On-time rate by week
  • Late deliveries by weekday
  • Cost per delivery

Marketplace sales

3 charts
  • Sales by month
  • Top products
  • Marketplace vs own store

About dashboards

How to connect

Connect CSV and file uploads, step by step.

  1. 1

    Open Explore

    Select where the file should go: usually your Orcabase warehouse.

  2. 2

    Load the file

    Click Load data and choose the file. Orcabase suggests a name from the file name.

  3. 3

    Name it

    Keep the suggested name or pick your own, like courier_august, and click Load data.

  4. 4

    Check it, then ask

    Explore opens the new data so you can check a few rows. Then ask the agent about it.

What you’ll need

  • The file: CSV, TSV, Parquet or JSON, up to 200 MB.
  • An Orcabase warehouse, or a database of yours that Orcabase is allowed to add tables to.
At a glance
How it connects
File upload
How fresh
A snapshot. Upload a newer file whenever you want fresh numbers.
Where your data lives
A new table in your Orcabase warehouse, or in a database you connect.
Who can set it up
Anyone in your organization.
Your login
Nothing to sign in to. You pick the file.
Cost
Included on every plan. Files can be up to 200 MB.
Upload filesStep-by-step, in the docs.

Questions

About CSV and file uploads and Orcabase

More in the docs: Upload files

Can I upload Excel files?
Save the sheet as CSV first: Orcabase reads CSV, TSV, Parquet and JSON. For a sheet you update often, keep it in Google Sheets and sync it instead.
How do I update an upload with a newer export?
An upload is a snapshot. Upload the new file under a new name, or remove the old one first. For data that changes often, a sync keeps it fresh for you.
Where should uploads go: my Orcabase warehouse or my own database?
Your Orcabase warehouse, if you have one: column types are detected there. Uploads into your own Postgres database come in as text, and need a login that can create tables.

Your CSV and file uploads data already has the answers.

Connect it, ask your first question, and pin the answers you want to see every week.