Dashboard
🔒 Read-only synthetic data
🔒
Read-only demo. Every control here is live — reassign a user, remap a project space, change node pricing or the allocation key and the whole model recalculates in your browser using the same cost engine as the Streamlit app. But nothing is saved and nothing is shared: changes exist only in this browser tab, Save buttons are disabled, and reloading restores the defaults. Explore freely; you cannot break it or affect anyone else.
✕

📊 Dashboard

Platform cost allocation across business units. Total cost is contracted and fixed — chargeback decides its distribution.
📈 Cost & Trend
⚡ Activity

Cost allocation trend by business unit

stacked, 7 months

Share of platform cost

Cost composition

dedicated / shared / overhead

Execution volume & failures

all business units

Execution minutes by Snaplex

🏢 BU Management

Define business units, cost centres and owners. User counts are derived from assignments in Asset Mapping — they are not typed in here.

Allocation detail

Edit business units

edits apply to this tab only

Cost vs execution minutes

bubble size = assigned users

Effective rate

cost per execution minute, log scale

🔌 Asset Mapping

Map Snaplexes to business units, and configure project-path and user allocation rules.
⚡ Snaplexes
📁 Project Mappings
👤 User Mappings
☁️ Import from Platform
Define each Snaplex, its type, and which BU owns it (for dedicated) or how its cost is shared. Dedicated Snaplexes are charged in full to their owner; shared and Cloudplex capacity is split by each unit's share of adjusted execution minutes on that Snaplex.

Snaplex inventory

Execution minutes by Snaplex

Monthly cost by Snaplex

nodes × cost per node
Project spaces are matched to business units by longest path prefix. This rule carries most of the attribution, because scheduled and triggered runs often have no requesting user. Spaces with no mapping fall through to Other / External.

Project space → business unit

Projects within a space

Each project inherits its space's business unit unless you assign it explicitly. An explicit assignment is a longer path prefix, so it wins — the same longest-prefix rule the engine uses. Per-project execution figures are apportioned from the BU-level fixture for the demo; the totals reconcile exactly to it.
Map platform users to business units. User mappings drive each BU's headcount, which in turn drives its share of platform overhead — so reassigning someone here moves real money on the Dashboard. Users marked Training are excluded from cost allocation and from headcount.

Domain rules

User → BU mappings

How the user and project lists get into the console.

From the Platform API

unavailable here
This page is static — no server, and no credentials. Anything that talks to a SnapLogic org lives in the Streamlit app. Credentials are deliberately absent from a publicly readable page.

From a users CSV

works in-browser
A user export from the platform admin console can be loaded here. It is parsed in your browser and never uploaded anywhere. Expected columns: Name, Email, Role, App access, Groups, Date added.

⚙️ Cost Configuration

Configure platform overhead costs, Snaplex node pricing, and the overhead allocation key.
💸 Platform Overhead
🖥️ Snaplex Pricing
🔑 Allocation Key
👁️ Impact Preview

Platform overhead

—
—
Startup overhead is added to every execution before the shared-Snaplex split, so thousands of sub-second runs are not treated as free next to a handful of long ones. Raise it and short-run-heavy units pick up more cost.

Overhead vs Snaplex cost

Node pricing per Snaplex. Selecting a node type fills in the standard rate; override it for contract-specific pricing.

Snaplex pricing

Monthly cost by Snaplex

nodes × cost per node

Overhead allocation key

Headcount comes from user mappings, not a typed-in number. Change who is assigned to what in Asset Mapping → User Mappings and the headcount side of this key moves with it.

Overhead split, current key

What your changes did, and what the other allocation keys would do instead.

Change vs bundled defaults

Effect of allocation key

same month, same total, four answers
The choice of key moves real money between units, so it should be agreed once and written down — not revisited per invoice.

⚖️ Allocation Model

How a pipeline execution becomes a line on a business unit's invoice.

Attribution

1

Drop excluded users

Training accounts, bots and leavers are removed before any cost is allocated, rather than charged to a catch-all.

2

Match project path

Longest matching /org/space prefix decides the business unit. This carries most of the load, because scheduled and triggered runs often have no requesting user.

3

Fall back

Anything still unmatched lands in Other / External so it stays visible instead of quietly disappearing.

Two paths, one caveat. On the live Platform API path an exact user-to-unit mapping takes priority over the project path, and unmatched executions are dropped rather than bucketed. The same month can therefore total differently depending on the data source.

Current inputs

live

Derived headcount

from user mappings

📄 Chargeback Report

Per-unit statement for the selected month — the artefact that gets sent to cost-centre owners.

Chargeback statement

3-month projection

linear trend on the last 3 months of usage