Skip to content

PointSav Documentation

The engineering library for the PointSav platform — operating systems and services for regulated businesses that own their data, their AI, and their record-keeping outright. Where the monorepo holds the code, this wiki holds the reasoning: architecture, services, security, and the governance commitments that bind future development.

PointSav Digital Systems

SYS-ADR-07: zero AI in Ring 1

← All revisions

cc603988 · PointSav Digital Systems ·

editorial(governance): fix adr-07-zero-ai-in-ring-1's service-input/service-extraction Ring misattribution (Track-B) — confirmed service-input has zero PDF/DOCX/XLSX deps (real Cargo.toml), is a deterministic migration+calibration tool (F1/precision/recall vs a labeled reference set, no model); confirmed real parser crates (lopdf/docx-rs/calamine, no pulldown-cmark, Markdown read as plain text) live in service-extraction (Ring 2, per this article's own table) not service-input (Ring 1) — architecturally significant for an article about the Ring1/Ring2 AI boundary; de-narrated the 2026-08-02 correction per register-documentation.yaml, fixed the ES pair which had never been corrected at all; also fixed a pre-existing duplicate body H1 (content-contract violation) on both EN+ES

View the full record as of this revision →

@@ -17,8 +17,6 @@ last_edited: 2026-07-11
cites: []
---

# SYS-ADR-07: Zero AI in Ring 1

SYS-ADR-07 is an architectural decision that prohibits AI inference from Ring 1
boundary-ingest services. It is not a preference or a guideline — it is a
hard constraint that all four Ring 1 services implement as code, with no
@@ -38,7 +36,7 @@ storage. In the current implementation, Ring 1 consists of four services:
| `service-fs` | [[worm-ledger-architecture|WORM immutable ledger]] — all Ring 1 writes land here |
| `service-people` | Identity ingest — Person, Anchor, and Claim records |
| `service-email` | Email ingest via EWS (Exchange Web Services SOAP) |
| `service-input` | Document ingest — PDF, Markdown, DOCX, XLSX |
| `service-input` | Reference-archive ingest — migrates and calibration-scores content into `service-fs` |

These services are the first point of contact for data entering the system.
Everything they write to `service-fs` becomes part of the permanent, auditable
@@ -155,18 +153,21 @@ from the WORM ledger — never inline during ingest.

### service-input

**Correction (2026-08-02, verified against canonical `origin/main`, not just this archive's local checkout):** this whole section misattributes document parsing to the wrong Ring. The real `service-input` (Ring 1, per this article's own §1 table) has zero PDF/Markdown/DOCX/XLSX dependencies in its `Cargo.toml` — it's a batch-migration/calibration ingest tool. The four parsers below actually live in `service-extraction`, which this article's own §1 table names as a **Ring 2** service — architecturally significant for an article whose thesis is the Ring-1/Ring-2 AI boundary. The specific crate names are also wrong: real deps are `lopdf` (not `oxidize-pdf`) and `docx-rs` (not `docx-rust`); no `pulldown-cmark` dependency exists anywhere in the corpus. `calamine` is the one accurate match. **Flagged, not resolved.**

`service-input` parses documents into text content:

- **PDF** — structural parsing via `oxidize-pdf`; extracts text spans
- **Markdown** — event-stream parsing via `pulldown-cmark`; strips HTML tags
- **DOCX** — ZIP + XML paragraph extraction via `docx-rust`
- **XLSX** — all-sheets tab-delimited extraction via `calamine`

In every case, the parser reads format-defined structure and emits text.
No semantic interpretation, no summarisation, no classification. The raw
extracted text is what lands in `service-fs`.
`service-input` migrates reference-archive content into `service-fs` and
calibration-scores the entity extraction Ring 2 later produces from it — it
does not parse documents itself. Migration and calibration are both
deterministic: migration copies content byte-for-byte into the ledger, and
calibration computes precision, recall, and F1 by comparing Ring 2's output
against a pre-labeled reference set. Neither step runs a model. A low F1
score flags an extraction-quality issue downstream — it never triggers
`service-input` to reinterpret or re-score anything itself.

Document parsing — PDF, DOCX, XLSX, and plain-text Markdown — happens in
`service-extraction`, a Ring 2 service (§5, below). That distinction matters
for this article's thesis: the parsers are
deterministic (`lopdf` for PDF, `docx-rs` for DOCX, `calamine` for XLSX;
Markdown is read as plain text, no dedicated parser), but they run one Ring
below the boundary this ADR governs, not at it.

---
Important Information

Corporate structure. PointSav Digital Systems ("PointSav") is currently a trade name of Woodfine Capital Projects Inc. ("Woodfine"), planned to become a wholly-owned Woodfine subsidiary upon incorporation. PointSav does not itself offer, sell, or solicit any security. Any securities offering associated with Woodfine's real-property direct-hold solutions is made exclusively by Woodfine, and only by means of the applicable Private Placement Memorandum.

No investment advice. This wiki's content is provided for engineering, operational, research, and development purposes. Nothing on this wiki constitutes investment advice or a solicitation to invest in any Woodfine partnership or direct-hold solution.

Intellectual property. The PointSav name, trade name, wordmark, and marks, together with all current and future PointSav- and Totebox-branded products, services, and offerings — and the software, source code, documentation, design system, and all related materials — are proprietary to Woodfine and its affiliates, except for components identified as open source. No rights are granted except as expressly set out in a written license or agreement. The full trademark notice appears in the footer of every page on this site.

Open source components. Portions of the platform are made available under permissive open-source licenses identified in the accompanying repository. Use of those components is governed by their respective license terms.

No warranty; informational use. Content on this wiki is provided for general informational purposes only and does not constitute a representation, warranty, or commitment with respect to product functionality, availability, pricing, or roadmap. Some articles describe planned or intended features, capabilities, and milestones — language such as "planned," "intended," "targeted," "may," and "expected" marks this forward-looking content, which is subject to change and does not constitute a commitment regarding future performance.

Confidentiality. Where an article describes an operational or deployment detail that is not intended for public disclosure, that article is not published on this wiki. Content here is general-purpose engineering documentation, not customer-specific configuration.

Jurisdiction. Woodfine Capital Projects Inc. is organized in British Columbia, Canada. References to the Sovereign Data Foundation on this wiki describe a planned or intended initiative only, not a current equity holder or active governance body.

Changes to this notice. PointSav may update this notice from time to time; the version posted on this page governs.

Not a filing system. This wiki is not a securities filing system, an electronic disclosure repository, or a substitute for SEDAR+ or any other regulatory filing system. Formal securities filings are made through the applicable regulatory filing system, not through this wiki.

Full disclaimer. This notice supplements, and does not replace, the full Disclaimers article. In the event of any conflict, the full Disclaimers article governs.

Read the full disclaimer →