Database Uploads allow you to bring data that's been processed outside of Logikcull into the application for search and review. Often, this will be data produced to you by another party, such as a legal document production or data processed by a vendor. With Database Imports, you can pull these already processed documents into Logikcull using special files, called load files, which contain metadata and file location information about the documents in the collection.
To upload your database import:
Go to the Uploads tab, click the "+" New button, and select "Production Upload" from the drop-down list.
2. Select "I'd like to import a data load" and checkmark Images, Natives and/or Text, depending on what was provided by the producing party. Usually, the production should have subfolders titled "Images", "Natives" and/or "Text" that will let you know what was provided.
3. Prior to zipping up the volume, find the load files (usually contained in a folder named DATA). Drag and drop, or click to upload your load files. The top box is for .DAT or .CSV files; the bottom box is for .OPT or .LFP files.
Helpful Tip: If the images are in multi-page PDF/TIFF format and the filenames are named by the bates number, attaching an image load file (i.e. OPT/LFP file) is optional.
4. Map your import fields. Logikcull will automatically map suggested matches based on the similarity of the field names, and mark the rows in green. It is recommended to check that they are mapped correctly; the definition of the field and the sample data that pulls from the first three lines of your load file can help as references to confirm or modify this choice. All fields will have to be mapped before you can move on to the next step. When mapping any fields, you can either choose to: (1) import them into an existing Logikcull Document field, (2) import them as a new custom field, or (3) not import them.
💡 Helpful Tips
If there is no corresponding Logikcull field for one of the imported fields, it may be helpful to choose the "import as a new field" option so the metadata is still searchable in Logikcull, as needed.
Mapping templates can be re-used when uploading productions from the same producing party, as the fields are expected to be consistent. If you already have one saved from a prior database upload that you'd like to use, you can load this from the dropdown at the top right. If not, it is recommended that you save a mapping template via the option at the bottom left.
If you have a large number of fields, you can checkmark select fields, or use the "bulk map" dropdown to quickly select unmapped/mapped fields, or deselect all fields. Then, you can choose to "import as new", "do not import", or "reset to Logikcull selections" in a single action.
5. Zip up the production volume (DATA, IMAGES, NATIVES, TEXT folders), making sure the folder structure is exactly as outlined in the load file. Upload the data volume. This should be the entire set of images/natives/text files provided as a compressed file with .zip extension. Input a name for this import. You can checkmark the option to attempt OCR on documents if no text is detected, or render an available native if no image is detected, so that Logikcull can render images and make your documents searchable.
If the archive is password protected, you must remove the password prior to uploading into Logikcull, then re-zip so that you upload the unprotected version of the volume.
For productions that include data sets with multi-page images in PDF or TIFF that are each named by their corresponding bates (begin doc) number, e.g. BATES000001.pdf or BATES0000001.jpg, you can skip loading an image reference load file OPT or LFP. Logikcull will automatically locate the images and load accordingly (i.e. common with Relativity productions).
NOTE: If you would like to use the text given by the producing party to search by, you will want to make sure that you mapped a "text file link" during step 2. With this option, the searchable text may be more reliable, but searching on the actual image itself for keywords depends on if the document have a text layer prior to upload. If not, you will have to use the "text" view of the document to scroll through any text hits. If you would like to search for keywords on the actual image, you would want to not import a "text file link", and make sure you choose the option to "attempt OCR if no text detected". However, please note that this may cause the upload to take longer to complete, and the size of your upload may expand more than expected due to the processing needed to be done. Moreover, text files can technically be loaded without a "text file link" in the metadata file, as long as the text files themselves are named with their corresponding bates (begin doc) number, e.g. BATES000001.txt. In addition, Logikcull will automatically locate and load searchable text in these cases.
If you run into any errors or have any questions about this workflow, feel free to contact our support team through in-app chat!