Skip to main content
Upload a CSV, Excel, JSON, or Parquet file and it instantly becomes a queryable dataset. No configuration, no schema definition — just drop the file and start asking questions.

​
Via the UI

Drag a data file into the chat or use the upload button. Zark detects the file type, processes it, and confirms when it’s ready:
  1. Drop a file into the chat
  2. Wait for processing (usually seconds)
  3. Ask a question about the data
What's the average order value by region?

​
Via the API

Upload a data file, then reference it in a chat request: 
import requests

# Step 1: Upload the file
with open("sales_data.csv", "rb") as f:
    upload_response = requests.post(
        "https://api.zarkai.xyz/v1/storage/files",
        headers={"x-api-key": "your_api_key"},
        files={"file": ("sales_data.csv", f, "text/csv")},
    )

file_id = upload_response.json()["fileId"]

# Step 2: Query it
response = requests.post(
    "https://api.zarkai.xyz/v1/chat",
    headers={
        "x-api-key": "your_api_key",
        "Content-Type": "application/json",
    },
    json={
        "messages": [
            {"role": "user", "content": "What's the total revenue by product category?"}
        ],
        "workspace_id": "wks-xxxxx",
        "file_ids": [file_id],
        "tools": ["database"],
        "tool_choice": "auto",
    },
)

print(response.json()["response"])

​
Multiple Data Files

Upload multiple related files and query across them: 
{
  "messages": [
    {"role": "user", "content": "Compare Q3 actuals to the Q3 forecast"}
  ],
  "workspace_id": "wks-xxxxx",
  "file_ids": ["file-actuals-123", "file-forecast-456"],
  "tools": ["database"],
  "tool_choice": "auto"
}
Zark creates separate tables for each file and can join or compare them based on your question.

​
Data Tables in Your Space

Once processed, data files become tables in your Space. These tables persist across conversations — you can query the same dataset in different chat sessions without re-uploading. Tables created from uploads are accessible to all tools:

​
Processing Status

Large data files show processing status during upload. You can check status via chat: 
Is my file done processing?

Show me the schema for the uploaded data
Once processing is complete, the full dataset is queryable.