Most annotation tools have one layer: the one you can see. Flowmono’s Freehand Tool has six. Here is why that matters.

Why Architecture Determines Defensibility

When a lawyer marks up a contract with a pen, the marks are visible. Whether they are defensible is a separate question, determined by context: who held the pen, when the marks were made, and whether the document being marked was the definitive version.

When an executive annotates a document using the Flowmono Freehand Tool, the marks are equally visible. But defensibility is not a separate question. It is built into the architecture that processes every stroke.

Understanding why requires looking at each of the six layers and what it contributes to the overall system.

Layer 1: PDF Layer

The PDF layer is the base document. All Freehand interactions occur on top of it, not inside it. The underlying PDF is never modified by the annotation process. This is the first line of document integrity protection.

When a document is annotated in many conventional tools, the annotations are written into the PDF itself. The original document and the annotated version become the same file. If an annotation is later disputed, there is no clean way to present the unannotated original because it no longer exists as a separate artefact.

In the Flowmono architecture, the PDF layer and the annotation layer are separate. The original document remains intact throughout the annotation process. The annotations are applied on top of it visually but stored independently.

The integrity of the original document is preserved by design, not by practice. No one needs to remember to preserve the original because the architecture does it automatically.

Layer 2: Annotation Layer

The annotation layer is where every signature, initial, text mark, and freehand drawing is stored. Because it sits above the PDF layer rather than inside it, annotations can be extracted, reviewed, and presented independently while remaining visually integrated with the document.

For the Free Signer experience, this means that all marks placed freely across the document, regardless of which page or position, are stored in the annotation layer as structured data. Each mark has coordinates, a type classification, and a relationship to the document version it was placed on.

Layer 3: Gesture Engine

The gesture engine translates Apple Pencil input into annotation events at under 16 milliseconds of latency. This is the layer responsible for the physical quality of the signing experience.

The gesture engine distinguishes between intentional marks and incidental contact, handles pressure sensitivity to produce natural stroke variation, interprets tilt angle for realistic shading in freehand markup, and processes palm rejection to prevent accidental marks when the hand rests on the screen during signing.

Without a sophisticated gesture engine, the annotation layer would receive noisy, unintentional input alongside the deliberate marks. The gesture engine ensures that what reaches the annotation layer is precisely what the signer intended to place.

Layer 4: Audit Engine

The audit engine is the layer that transforms a signing experience into legal evidence. It operates at the stroke level, not at the session level.

For every discrete stroke placed by a Freehand Signer, the audit engine records: the exact timestamp (to the millisecond) at which the stroke was initiated and completed, the spatial coordinates of every point in the stroke path, the identity of the authenticated user placing the stroke, and the document version hash at the moment the stroke was placed. This record is stored in a tamper-evident structure that makes post-event modification detectable.

The audit engine’s output is not a summary of the signing session. It is a complete, chronologically ordered, cryptographically protected record of every annotation event. If the document is challenged five years after signing, the audit record can reconstruct exactly what was placed, by whom, where, and when.

Layer 5: Workflow Engine

The workflow engine handles what happens after the annotation is complete. In the context of the Free Signer, this means routing the annotated document to its next destination: the next signatory, an approval queue, an archive, or any other configured workflow step.

For complex multi-party signing scenarios, such as a construction project requiring signatures from a contractor, a site engineer, and a client representative, the workflow engine manages the sequence, tracks completion status, sends notifications, and ensures that the document does not proceed to the next step until all required interactions are complete.

The workflow engine also handles the Free Signer’s specific status within a larger workflow. When the Free Signer completes their interaction and submits, the engine registers their completion and triggers the next workflow event, without requiring any manual intervention from the initiator.

Layer 6: AI Intelligence Layer

The AI intelligence layer analyses the annotations placed in the annotation layer and adds analytical enrichment that makes the document’s signing history readable, searchable, and actionable.

1. Signature Detection
Identifies marks that constitute signatures within a broader set of annotations, isolating them for specific audit and attribution purposes.

2. Annotation Classification
Categorises each mark: signature, initials, text annotation, or freehand drawing. This classification is stored in the audit record alongside the raw annotation data.

3. Handwriting OCR
Converts handwritten text annotations into machine-readable, searchable data. A handwritten condition added by an executive during a Freehand signing session becomes a searchable data point in the document record.

4. Approval Intelligence
Infers the nature of an annotation from its content and context. A tick mark adjacent to a clause, combined with a signature, may constitute approval of that clause. The AI layer classifies this combination and records the inference in the audit trail.

5. Compliance Validation
Checks that the completed document meets configured signing requirements. For workflows that require a minimum set of annotation types, such as a signature plus initials on a specific page, the AI layer validates that the requirements have been met before allowing the document to proceed.

Why Six Layers and Not One

A single-layer annotation tool can record that a mark was placed on a document. It cannot simultaneously ensure document integrity (PDF layer), guarantee input quality (gesture engine), produce a legally defensible audit record (audit engine), route the result automatically (workflow engine), or make the result analytically useful (AI intelligence layer).

The six-layer architecture is what makes the Flowmono Freehand Tool appropriate for enterprise document governance rather than just useful as a signing convenience. Each layer contributes a property that the other five do not provide. Together, they produce an annotation environment where every mark is visible, governed, defensible, and intelligent.

Explore the full six-layer Freehand Tool architecture at www.flowmono.com.