Skip to content

Uploading documents to the Lookup collection

You can upload different types of documents to the Lookup Collection in the OneReach.ai system, including .csv, .pdf, .html, and .txt files.

Accepted document types:

  1. text/csv
  2. application/pdf
  3. text/html
  4. text/plain

Each of these document types has its own specific format, but they all contribute to enhancing the comprehensiveness and utility of your Lookup collection.

CSV document format

CSV (Comma-Separated Values) files are simple text files with a structured format. They store tabular data in a plain-text form. For Lookup collections, it's notable that the field Content is always required, as it includes the searchable passages. Any other fields in the CSV will be regarded as custom metadata, or "properties", of those passages.

CSV file structure

Arrange the CSV file in such a way that the first row consists of the column names. The "content" column, which is critical, followed by any additional columns that represent properties.

Example:

content, iconUrl, docId
"My first passage content", "http://mysite.com/icon1.png", "doc1"
"My second passage content", "http://mysite.com/icon2.png", "doc2"
content, iconUrl, docId
"My first passage content", "http://mysite.com/icon1.png", "doc1"
"My second passage content", "http://mysite.com/icon2.png", "doc2"

In this example, content is a required field, and iconUrl and docId are properties that provide additional information about the passages.

Uploading a CSV and other formats

Follow these steps for uploading a document (in any of the above format) to a collection:

  1. Open the Lookup UI in OneReach.ai.
  2. Create a new collection or select an existing one.
  3. Click Add Document.
  4. Provide a URL or choose a file from your computer for the document that you want to upload. Make sure the document is in the CSV format as described above (or in any of the other accepted format).
  5. If you've defined additional fields in your CSV other than content, make sure these fields are set up as properties in your collection.
  6. Click Properties and then Add properties, entering the name and datatype for each additional field as required.
  7. Click Upload.
  8. The system processes the document and breaks down the text in the content field into passages and assigns corresponding property data from other fields. This operation may take some time depending on the size of the document. These passages now form part of your Lookup collection.

NOTE

When uploading PDF, HTML, and Plain Text documents, make sure they are well-structured and organized to optimize their integration into the Lookup collection. There are no strict format rules here, but a coherent and logically structured document helps the system to build more accurate and helpful searchable passages.