Spreadsheets are not documents made of paragraphs. Sending a workbook through the document parser produces hundreds of chunks whose text is cell | cell | cell, which pollutes semantic search and still cannot answer “what is the total COGM for B000”. Pinter therefore treats tabular files (csv, xlsx, xls) as data, not text:
This is deliberate: retrieval over spreadsheet rows returns plausible-looking noise, while a GROUP BY returns the exact number. Tabular documents are excluded from semantic search.

What happens on upload

1

Validate and store the raw file

The extension must be csv, xlsx, or xls, and the content must match the extension. The original file is stored unchanged at knowledge/<asset_id>/v1_<filename>.
2

Skip the document parser

No parser job is submitted and no chunks or embeddings are created. The version is marked ready with chunk_count = 0, so the document appears as processed right away.
3

Normalize on first access

The first time the document is opened — in the UI detail view or through an agent tool or route — the backend downloads the raw file and builds one SQL table per sheet. The result is cached in GCS as knowledge/<asset_id>/tabular/<checksum>-v1.sqlite.gz, so later requests on any replica skip the work.
Normalization for a typical report takes well under a second; a 72-sheet workbook takes about 13 seconds on first access only.

Normalization rules

Each worksheet becomes its own table, named after the sheet: A workbook with a single sheet, and every CSV file, uses the table name data. Column names are always quoted identifiers because real headers contain spaces and parentheses ("Stok Admin (Ton)", "Beban Administrasi dan Umum"). The normalizer then applies these steps, recording every decision in the artifact metadata:
Limits per document: 25 MB file, 500,000 rows, 1,000,000 cells, 500 columns, 100 tables. Sheets that exceed a limit are skipped with a warning when the workbook has other usable sheets; a single oversized sheet fails the document with a clear message.

Read tables as an agent

Agents never write SQL. They send a structured query — table, columns, filters, aggregations — and the backend compiles it with validated identifiers and bound parameters. That removes SQL injection entirely and keeps the model from inventing functions or joins that do not exist. Endpoints live under the knowledge document path and require the caller to have use permission on the document.
GET /tables returns the schema the agent needs before querying:
Only the selected table carries its column list and value facets; the others are listed by name and size. This keeps a 72-sheet workbook from flooding the model context (a five-sheet workbook costs ~750 tokens instead of ~2,500). values holds up to 20 distinct values for up to five low-cardinality TEXT columns — the agent filters on real values ("ANPER": "B000") instead of guessing the spelling. sample is columnar so column names are not repeated per row.

Query parts

Supported op values: eq, neq, gt, gte, lt, lte, in, not_in, contains, starts_with, between, is_null, is_not_null. Supported fn values: count, count_distinct, sum, avg, min, max. sum and avg are rejected on TEXT columns instead of returning a wrong number.
The response is the same shape as the row preview:
truncated: true means more rows matched than limit allowed — raise limit or paginate with offset instead of reporting the partial count as the total.
In chat, agents use the built-in tools get_table_schema(document_id, table?) and query_table(document_id, table?, columns?, filters?, group_by?, aggregations?, order_by?, limit?). Same builder, same validation.
POST …/tables/query still accepts raw SQL for analysts and integration tests that must express something the builder cannot (e.g. CASE, window functions). It is authenticated and runs read-only, but it is not offered to agents — keep it out of prompts and automations driven by a model.

Compare raw and parsed data in the UI

The document detail view shows both representations, so you can confirm what ingestion changed:
  • Parsed — the normalized tables (header detected, subtotals removed, typed columns). When a workbook has several sheets, a selector switches between them.
  • Raw file — the untouched grid, columns labelled A, B, C, including the title rows. Useful when a header was detected in the wrong place.

Limits and errors

Known limitations

  • Sheets that stack several unrelated tables vertically are read as one table; only the first header row is used.
  • Duplicate column labels in the source get _2, _3 suffixes when there is no group row above them to disambiguate.
  • Dates are read as text; cast them in SQL when you need date arithmetic.
  • Cells are read from the cached formula results stored in the workbook. A workbook that was never recalculated may show stale or empty values.