Skip to content
View Autophuwit's full-sized avatar
  • TRUE CORPORATION

Block or report Autophuwit

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Autophuwit/README.md

Hi, I'm a data analyst & data scientist in telecom

I turn customer, sales and network data into decisions that move revenue — building the data, analysis and models behind them, and following through until the business acts on the results.

  • Telecom, Thailand — mobile, home broadband and retail channels, in Customer Value Management
  • End to end — data engineering (ETL, star schemas, SQL on Vertica / Netezza / SQL Server), analysis, dashboards, and ML when it earns its place
  • Business first — every project starts from a business question and ends with a number that changed

Featured projects

These are real work projects. Figures are rescaled or synthetic to protect confidential data; the patterns, rankings and conclusions are unchanged.

Data engineering & analytics

Project What it does Impact
Sales Monitoring Dashboard · ▶ live demo Python ETL → SQL Server star schema → in-browser metric engine. 3 dashboards (desktop + mobile) for 7-11 and retail channels, click-to-filter, tests and CI Drove 5 sales game plans: +10% to +15% growth in focus stores
Digital Product Analytics MAU, ARPU uplift and cohort-survival reporting for digital products; root-cause deep-dive on why OTT buyers churn Sales guardrails predicted to increase OTT customer retention by 3%
5G Customer Analytics 5G flags in the High Value report with point-in-time joins; provisioning framework joining subscribers to their most-used cell site Prioritised customer groups and areas: who to provision first, where to build, where 5G fails
Home Broadband Device Analytics Turns router Wi-Fi logs into a household device profile for ~2.5M homes and finds the upsell gaps 3 sized upsell use cases (Mesh Wi-Fi, CCTV, TV box), 575K–730K homes each

Machine learning

Project What it does Result
Telecom Alarm Grouping (GNN) · ▶ live demo Groups network alarms into incidents with a GATv2 graph neural network trained on weak labels, average-linkage clustering, a live 5-minute engine and a 3-D incident view 12.1% of tickets merged into incidents at 81.7% precision

Toolbox

Data & SQL — Vertica · Netezza · SQL Server / T-SQL · star-schema modelling · ETL design Python — pandas · data pipelines · graph neural networks (GATv2) · clustering · unit tests Visualisation — JavaScript dashboards (Chart.js, Leaflet) · interactive 3-D views · PowerPoint for management Delivery — Git & GitHub Actions · GitHub Pages · Cloudflare

Popular repositories Loading

  1. telecom-alarm-grouping-gnn telecom-alarm-grouping-gnn Public

    Grouping telecom network alarms into incidents with a graph neural network (GATv2) trained on weak labels, plus average-linkage clustering, a live 5-minute engine with stable incident IDs, a 3-D in…

    Jupyter Notebook

  2. sales-monitoring-dashboard sales-monitoring-dashboard Public

    Sales monitoring dashboards for Modern Trade (7-11) and Retail channels: Python ETL, SQL Server star schema, in-browser metric engine, desktop + mobile. Live demo on synthetic data.

    Python

  3. 5g-customer-analytics 5g-customer-analytics Public

    5G customer analytics: 5G flags added to a High Value churn report, and a provisioning-prioritisation framework using cell-site coverage - who to provision, where to build, where 5G fails.

  4. home-broadband-device-analytics home-broadband-device-analytics Public

    Household device analytics from router Wi-Fi logs: a ~400-rule SQL classifier for device group & brand, device-combination analysis, and three sized upsell use cases (Mesh Wi-Fi, CCTV, TV box).

  5. telco-digital-product-analytics telco-digital-product-analytics Public

    Telecom digital-product analytics: MAU, ARPU uplift and cohort survival reporting, plus an OTT churn root-cause deep-dive whose sales guardrails close the retention gap. SQL on Vertica & Netezza.

  6. Autophuwit Autophuwit Public

    Profile README