Program Manager — Tower Editor

Reinsurance programmes.
Structured. Visual. Documented.

As the Program Manager, you design the reinsurance tower structure — the chain of protections a cedant purchases, layer by layer. You build the architecture that makes every programme legible and auditable.

🏗️ Programs & towers

🕸️ DAG-based structure

⬡ Node types

→ Link types

🧩 Dimensions

' . "\n"; $content_html .= $sec['html'] . "\n"; $content_html .= '

Key concepts

Three levels — one hierarchy.

Level	Object	Description	Example
1	Programme	Top-level container grouping all towers for a given cedant or risk portfolio.	Standard
2	Tower	A single protection layer structure — a directed graph encoding how losses flow from gross to net.	QS + XL Risk
3	Node	A position in the flow — gross loss, ceded amount, net retention. Each has a type and optional dimension annotations.	Gross loss, Net retention

💡

The tower is a DAG. Directed, Acyclic Graph — flows always go from gross input to net output, never backwards. The platform detects and flags any accidental cycle.

Sidebar

The left sidebar lists all your programmes, each expanded on load to reveal its towers.

Navigation

Click a programme name to expand or collapse its tower list. Click a tower row to load it into the canvas. The active tower is highlighted in purple. The breadcrumb in the header shows Programme → Tower.

Search

Type in the search box to filter programmes and towers in real time. Click any result to navigate directly.

Creating objects

Use + New tower inside a programme, or + New program at the bottom of the sidebar. Both open a modal asking for the label and (for towers) the parent programme.

Building a tower

A tower encodes how losses travel through the reinsurance structure — from gross position through cessions and recoveries to the net retention.

💡

Think of it as a loss flow diagram. Gross loss at the top, net at the bottom. Each node is a position; each link qualifies how one position feeds into another.

Workflow

Open a tower

Click any tower row in the sidebar. The canvas loads its nodes and links. If no positions are saved, Auto-layout runs automatically.

Add nodes

Click + Node or double-click the canvas. Set label and node type in the inspector, then click Apply.

Draw links

Hover a node until the port appears at its bottom edge. Drag to another node. A link type picker opens — select the appropriate type.

Attach dimensions

Select a node → Dimensions tab → ＋ Add dimension. Choose from the catalogue. Click Apply.

Save

Click 💾 Save or press Ctrl+S. Nodes, links, positions and dimensions are written atomically. Cycle detection runs automatically.

⚠️

Save is all-or-nothing. Unsaved changes exist only in the browser. The header shows ● unsaved when there are pending changes.

Canvas controls

Action	How
Pan	Click and drag empty canvas
Zoom	Scroll wheel
Fit view	⊡ Fit button
Auto-layout	⬡ Auto-layout — recomputes positions top-to-bottom
Select node/link	Click it — opens inspector
Delete node	Select → ✕ badge, or `Delete`
Delete link	Select → 🗑 in inspector, or `Delete`
Quick add node	Double-click empty canvas
Save	`Ctrl+S` or 💾 Save

Inspector panel

Clicking a node or link opens the inspector — your editing workspace for that element.

Context	What you do here
Node — Properties	Set label and node type. Click Apply to confirm locally, then Save.
Node — Dimensions	Add or remove dimension attachments. Apply then Save.
Link	Change the link type. Apply then Save.
Tower info (nothing selected)	Programme and tower stats, cycle flag, and the ✏️ Rename button.

💡

Apply vs Save. Apply updates the in-memory canvas. Save writes to the database. Always Save after all your Apply actions.

Node types

Node types define the role of each position in the loss flow. They are colour-coded on the canvas.

The platform ships with two seed types: LOSS (a loss position — gross, ceded, or net) and LOSS_COMMON_ACCOUNT (a common account loss shared between cedant and reinsurer). You can add as many custom types as your structures require via ⬡ Node types in the Reference section.

💡

Node types are fully customisable. Create types that match your vocabulary: PREMIUM, REINSTATEMENT, AGGREGATE_LIMIT, etc. Each type gets its own colour and is available immediately in the node inspector.

Link types

A link connects a parent node to a child node. Its type qualifies the nature of the flow.

Type	Meaning	Typical use
CESSION	A portion is transferred to the reinsurer	QS cession, XL cession, SL cession
ACCEPT	Incoming acceptance from another party	Facultative acceptance, pool participation
RETRO	Retrocession — a reinsurer cedes onward	Retrocession structures

💡

Link types are fully customisable. Add types matching your structures via → Link types in the Reference section. Each type gets its own arrowhead colour.

Dimensions

Dimensions attach documentary axes to nodes — they specify what must be described at each position in the tower.

Seed category	Dimensions	Purpose
FLOW — Flow State	GROSS · RETENTION · RECOVERY	Position in the loss flow chain
COVERAGE — Coverage Type	QS · XL_RISK · XL_EVENT · XL_COMMON · SL	Type of reinsurance coverage at this node

Attaching dimensions

Select the node

Click the node on the canvas.

Open Dimensions tab

Click the Dimensions tab in the inspector.

Add, Apply, Save

Click ＋ Add dimension, choose from the dropdown (grouped by category), repeat as needed. Click Apply then Save.

Managing the catalogue

Open 🧩 Dimensions in the Reference section. Use + Add category and + Dim. to create, or the Edit / Del buttons to modify. A dimension cannot be deleted while attached to a node.

Managing node & link types

The Reference section gives you full CRUD control — code, label, colour, and display order for each type.

⚠️

Codes are immutable after creation. A type in use cannot be deleted. Rename labels freely — the code is the stable identifier stored on each node or link in the database.

The colour picker assigns the rendering colour for nodes on the canvas and arrowheads on links. Choose contrasting colours for types that frequently appear together in the same tower.

Data Checks

Live queries on your tenant data — accessible from the Analytics section of the sidebar.

View	What it shows
Inventaire des towers	All towers with node count, link count, and cycle flag — programme by programme.
Nœuds sans dimension	Nodes with no dimension attached — useful before a programme review.
Towers avec cycle détecté	Towers flagged with a structural cycle — must be corrected before signoff.

Click ▶ Run on any view, then ⬇ Download CSV to export. The toolbar ⬇ CSV exports the current tower's nodes and links as a flat file.

Database purge

Three cleanup operations for structural orphans. Always diagnose first — it is always safe and shows exactly what will be removed.

⚠️

All purge operations are irreversible. Review the diagnostic list before confirming.

Code	What it cleans	Typical cause
P1	Nodes whose parent tower was deleted — cascades to their links and dimension attachments	Tower deleted outside normal flow
P2	Links whose parent or child node no longer exists	Node deleted without cascading its links
P3	Dimension attachments whose node or dimension no longer exists	Node or dimension deleted leaving stale attachments

Recommended order: P1 → P2 → P3.

Cycle detection

A tower must be acyclic — no node can be its own ancestor. RI-TOOL checks this on every save.

If a cycle is detected, an orange banner appears at the top of the canvas and a red dot appears on the tower row in the sidebar. The tower is still saved but is structurally invalid.

⚠️

To fix a cycle: find the link that feeds back into an ancestor and delete it. Resave — the banner disappears when the graph is acyclic.

⚡

New to the platform? Your workspace is already built.

Every new account includes a complete reference framework, a live Profit Commission example with 5 actuary graphs, and 2 SOA statement templates — included at signup.

→