Operations Intelligence Command Center

Portfolio-safe BI showcase for turning messy operational workflow signals into an executive command center: KPI health, backlog aging, automation impact, friction concentration and deterministic management actions.

This is product-agnostic BI, not a Power BI project. There is no .pbix, DAX model or Power Query artifact in this repo. The point is the analytical method and delivery quality: deterministic synthetic data, SQL metric lineage, reproducible validation and a self-contained dashboard that opens offline.

60-Second Review Path

1. Open dashboard/index.html for the finished command center.
2. Scan the Decision Brief, Filtered executive focus, Friction Heatmap and Risk Pareto sections.
3. Check sql/ for the metric queries and operations_bi/validate.py for quality gates.
4. Read docs/kpi-definitions.md if you want the formulas behind the cards.

Business Problem

Transformation and operations leaders often have workflow data, but not a clear answer to: what is late, why is it late, where is manual work hurting outcomes, and what should management fix first?

This repository models that operating-review problem using synthetic workflow records across procurement-to-payment, customer onboarding, service requests, inventory replenishment and management reporting.

Executive Decision Questions

Question	Where to inspect
Where are SLA misses and rework concentrated?	Friction heatmap, sql/08_friction_heatmap.sql
Is automation improving cycle time and quality?	Manual vs Form/API outcomes, sql/07_channel_comparison.sql
Which risk segments deserve management attention first?	Risk Pareto, sql/09_pareto_risk.sql
Which open items need follow-up now?	At-risk queue, sql/05_follow_up.sql
What actions should leadership take?	Decision Brief generated in operations_bi/pipeline.py

Key Findings From The Synthetic Snapshot

Current snapshot: June 2026, generated from 749 synthetic workflow records as of 2026-07-10.

Finding	Synthetic evidence
SLA performance is below target but improving.	June SLA hit rate is 53%, up 0.9 percentage points month over month, against an 82% target.
Automation is near the desired threshold.	June form/API share is 70%, up 5.7 percentage points month over month.
Manual intake remains the main quality drag.	Manual work averages 9.9 days, with 41% rework; API work averages 5.8 days, with 21% rework.
Backlog risk is still visible.	June has 22 open backlog items, all overdue, although backlog is down by 9 month over month.
Management should focus before adding controls.	Top 3 risk segments hold 22% of value-at-risk; the highest friction segment is Procurement / Customer onboarding.

All values above are synthetic and generated by this repository.

Architecture And Metric Lineage

flowchart LR
    Generator[Deterministic synthetic generator] --> CSV[data/synthetic_operations_events.csv]
    CSV --> SQLite[In-memory SQLite model]
    SQL[sql/*.sql metric queries] --> SQLite
    SQLite --> Snapshot[dashboard/snapshot.json]
    Snapshot --> Dashboard[dashboard/index.html]
    Snapshot --> Preview[docs/dashboard-preview.svg]
    Snapshot --> Actions[Deterministic management actions]
    Validate[Validation and tests] --> CSV
    Validate --> Snapshot
    Validate --> Dashboard

Loading

Metric path: source CSV → SQLite schema → SQL query → JSON snapshot → HTML/SVG dashboard. The generated dashboard does not fetch remote assets, call APIs or depend on a local server.

What This Proves

Capability	Evidence
Operations intelligence	Backlog aging, SLA risk, friction scoring, Pareto concentration and action ranking.
BI delivery	Source grain, KPI definitions, SQL lineage, target context and a polished dashboard surface.
Data quality	Deterministic generator, schema checks, reconciliation tests, dashboard section checks and CI.
Business translation	Management questions and actions are tied directly to explainable metrics.
Engineering hygiene	Standard-library Python, SQLite, generated artifacts, tests, CI, .gitignore and MIT license.

Quick Start

python -m operations_bi build
python -m operations_bi validate
python -m unittest discover -s tests

Then open dashboard/index.html in a browser.

Live Demo Readiness

The repo includes a root index.html entrypoint for GitHub Pages. Pages is not currently enabled on the public repository. To make the live route work, enable GitHub Pages from main / root in repository settings; the expected URL will be https://mypoorbrain.github.io/operations-intelligence-command-center/.

Repository Map

Path	Purpose
operations_bi/	Deterministic data generator, SQLite pipeline, dashboard renderer and validator.
data/	Synthetic workflow-event CSV generated from code.
sql/	Metric, trend, channel, friction, Pareto and follow-up queries.
dashboard/	Generated self-contained dashboard and JSON snapshot.
docs/	KPI definitions, architecture, data dictionary, generator assumptions and portfolio context.
tests/	Regression tests for data shape, KPI reconciliation, decision layers and output safety.

Privacy Boundary

This repository contains no employer data, client records, personal contact data, credentials, secrets or private operating records. All records are synthetic and generated deterministically.

Intentional Limits

This is a public portfolio artifact, not a connected production dashboard. A real deployment would need governed source ownership, access controls, row-level security, scheduled refresh, dashboard ownership and agreed KPI governance.

License

MIT. Synthetic data and code may be reused as reference patterns.