NUCLEUS
Executive Analytics & Decision Intelligence Platform
An independently developed analytics platform by Shemal Dave, PhD, PMP — demonstrating end-to-end ownership across governed metrics, data modelling, executive BI UX, forecasting, anomaly detection, player analytics, experiment design and AI-assisted querying.
No sign-up · no password · read-only · fully synthetic dataset
The full analytics stack, working
Every capability below is live in the demo — nothing is mocked in the UI. Click any dashboard and the numbers come from the same SQL metric layer.
Hold, RTP, rotation, net per store per day — each defined once in SQL. Dashboards and the AI read the same views, so there is exactly one version of every number.
Dense, boardroom-grade screens: CEO health, executive summary, trends, geography, leaderboards, game economics, player retention, promotion design.
Holt-Winters forecasts with confidence bands per location, and a severity-ranked exception list of what broke out of band this week.
Cohort retention triangles, behavioural segments and churn-risk scoring over a synthetic player base.
An 18-run Taguchi array for promotion design — factor effects ranked by signal-to-noise, with a recommended configuration.
Dr. D answers in plain English by writing read-only SQL against the governed views, with every question written to an audit log.
State, platform, product and period filters persist across every page; full dark/light theme support.
Row-level security, read-only query paths, server-mediated AI with the key never touching the browser.
A governed metric layer, end to end
Designed & built by Shemal Dave
Nucleus is an analytics platform I designed and built end-to-end as a portfolio project: the warehouse schema, the governed SQL metric layer, the forecasting engine, the executive UI, and the AI assistant that reads the same views the dashboards do.
The dataset is fully synthetic — a simulated distributed route gaming network — but every screen behaves the way a production deployment would, because every number is computed in SQL, never in the browser.
- PhD — applied research in statistical modelling and decision systems.
- PMP-certified program leadership across data platform builds.
- Nucleus is an independently developed portfolio build — designed and implemented from scratch to demonstrate analytics architecture, executive BI, statistical modelling, metric governance and AI orchestration skills.