Skip to main content
When you drop a file into your IP Vault, it doesn’t become instantly searchable. Pria takes it through five steps that read the text, cut it into searchable passages, clean them up, index them so they can be found by meaning, and mark the file ready. Knowing what each phase does — and what each status means — helps you spot trouble early and reprocess only what needs reprocessing.

​
The journey from upload to searchable

Each step saves its work before the next one begins, so if something goes wrong only that step is retried — a 200-page PDF doesn’t start over from scratch because the last stretch of indexing failed. You can watch the journey live on the file card in the vault. The progress bar advances through the phases, and the percentage shows real work done (not interpolated time).
Files uploaded with Skip Indexing checked bypass this pipeline entirely — they’re stored for safekeeping and download but never become searchable (and never incur indexing cost). Run Reprocess Content later if you change your mind.

​
The 5 phases

Pria reads the source bytes and turns them into clean text.The card sits between 5% and 40% during extract. Long videos and large PDFs spend most of their time here.
The extracted text is split into segments of about 2,000 characters each by default (roughly 300–400 words; a Digital Twin’s admins can pick 1,000 to 6,000 characters with Chunk Size (Search & Knowledge Graph) in Instance Settings › Knowledge). Segments overlap by about 20% so a sentence that lands on a boundary still appears whole in one of the two neighbours.Segment boundaries snap to the nearest paragraph or sentence end — chunks never tear mid-thought.The card runs 40% to 50% during chunking. Even huge files finish this phase in seconds.
Each passage gets a quick automatic check. Clean prose passes through untouched — no AI call, no cost. Passages that look noisy (scanning errors, broken formatting, leftover menus) are rewritten into cleaner prose by your Digital Twin’s summary model.Sanitize is enabled by default, but your administrator can turn it off per Digital Twin.The card runs 50% to 75% during sanitize. The bar advances in real time as each batch comes back from the model.
Every passage is turned into a numeric fingerprint of what it means (a vector embedding), so Pria can compare your question against all of them at once and find the closest matches — by meaning, not exact words.Pria optionally enriches each segment with 2–3 AI-generated keywords plus a document-level summary header before embedding. This makes single-segment matches more findable.The card runs 75% to 95% during embed.
The final phase is cheap closing work:
  1. Vault Health Grade is computed from the final segment set.
  2. The file is marked Included. That’s the flag Pria looks for — until it’s set, the file is skipped when Pria searches your documents.
  3. Scratch data is cleaned up.
  4. Token usage is rolled into the billing record.
Until the status flips to Included, retrieval will skip the file entirely. The card jumps from 95% to 100% during finalize.

​
Reading the progress bar

The bar on each file card shows the active phase and the percentage of work completed within the pipeline. The percentage reflects actual segments processed, not interpolated time — so 50% really means half the segments are through.
The counter updates in batches, not one segment at a time. On short files the bar may pause briefly before jumping ahead. That’s normal — work is still in flight.

​
Vault Health Grade

After finalize, every file gets a Vault Health Grade — a quick read on how usable the file is for retrieval. The badge appears on the file card and contributes to your overall Vault Health letter grade (A–F).

​
What affects the grade

The per-file score measures how much of the file’s extracted text is usable prose. Segments dominated by code, markup, or raw data (HTML/CSS/JavaScript, dense JSON, SQL, bracket-heavy text) count against it — they embed poorly and add noise to retrieval. The vault-level letter grade combines that signal with the operational health of your whole inventory: files in an error state, Included files with no indexed segments, missing source files, and files that never get retrieved all drag the grade down. See Vault Health for the full category breakdown and how to fix each one. If a file scores poorly, open it in the File Preview to see what’s flagged, then run Sanitize All on its segments or re-upload a cleaner source.

​
Common issues

​
Re-processing a file

Open the file’s action menu and pick Reprocess Content: Reprocessing keeps the file’s identity, so any links you shared keep working. See Reprocessing for details.

​
Confidential files

Confidential is a per-file setting for files in shared vaults (a Digital Twin’s or an account’s). When set:
  • Only you, the uploader, can open the file. Other members — including admins — can’t preview or download it; where its details would appear they see a ”🔒 CONFIDENTIAL” marker instead.
  • Processing is unchanged, but only for you. Confidential files go through the same pipeline and stay searchable in your own conversations; other members’ questions never draw on them — the Digital Twin’s search, knowledge graph and file tools all leave another person’s confidential file out.
  • For you, everything works normally — full citations, full preview, full retrieval in your own conversations.
You can toggle Confidential before upload (in the upload dialogue) or any time after (file action menu → Make Confidential / Remove Confidential). See What Confidential really means for the full picture.