Snapshot: I build analytics systems that cut reporting time, stabilise ops, and make promotions accountable.
0
End-to-end analytics projects
SQL · Python · BI
0
Reporting efficiency uplift
Automation focus
0
Domains: retail, ops, supply, print
Cross-industry
Short demo of the Retail & Logistics Intelligence Hub.
Retail & Logistics — walkthrough
End-to-end analytics hub for routing optimisation, forecasting, and operations insights.
Built using: SQL · Power BI · Azure · Python
Inspired by internal Microsoft reporting patterns (demo only).
Skills & Experience
Interactive skill map — click a node to filter projects. Drag to pan.
Tip: click a node to filter projects.
Selected skill:Click a node to see details.
Work · Project Gallery
Use search, personas, or tags to explore case studies.
Systems & Playbooks
Reusable analytics systems I keep reaching for.
Playbook
Promotion Profitability Engine
End-to-end framework that combines promo flags, uplift modelling, and margin tracking to answer
“Did this promotion actually pay off?” in one place.
Inputs: POS + campaign data · Outputs: ROI view, cannibalisation risk · Stack: Python, SQL, Power BI
System
Logistics Ops Healthboard
Route performance, SLAs, and exception management in a single hub designed around ops stand-ups,
not just pretty charts.
Inputs: telemetry, order logs · Outputs: daily health score · Stack: Snowflake, SQL, Tableau
Playbook
Anomaly Radar for Supply Events
Streaming anomaly detection pipeline with clear thresholds, alerting rules, and dashboards that
separate “noise” from “drop everything” events.
Inputs: event streams · Outputs: alert feed + triage board · Stack: Spark, Azure, Power BI
About · How I Work
Working principles
A few rules I use to keep analytics useful, not just “interesting”.
Make it measurable
Every dashboard must answer a decision question and have an owner. No orphan visuals.
Design for the meeting
I design views for weekly stand-ups and monthly reviews, not vague “exploration”.
Automate the boring
Manual Excel steps are red flags. If it repeats, it gets scripted or modelled.
Stay explainable
Models are only as good as the story stakeholders can tell with them in 2 minutes.
Timeline & credentials
2025 — Present · Data Analyst (Vendor) at Microsoft via BlitzPath Innovation.
2023 — 2024 · Data Analyst at ESVADHYAYA & consulting projects.
Before · Analytics across education, manufacturing, and ops.
Selected certifications
Machine Learning with Python — freeCodeCamp
Google Analytics (GA4) — Google Skillshop
SQL (Intermediate) — HackerRank
Programming Foundations with Python — CodeSignal / HackerRank
Notes from the field
I’m currently exploring how to bring promotion analytics, supply chain anomaly detection,
and geospatial delivery insights into one “ops command center” — with real-time-ish data and
less slideware.
Contact
Shoot a short note — demo submit only (replace with webhook in JS).