Software & automation
Shopskon: a small business that runs on automated, tested systems
A Lahore e-commerce brand built in ten days on one PC: static website, search and product feeds, an owner dashboard with weekly reports, and auto-publishing through official APIs.
Lead: Technology Lead · Software and workflow engineering project by our technology lead, 7–16 October 2026, built with Claude Code and free, locally run tools. Shopskon is an e-commerce business, not a research project; it is shown here as evidence of engineering method.
The problem
A one-person shop with no team, no advertising budget and unreliable power needed a fast, trustworthy store, visibility on search and social platforms, and a way to measure what works.
Approach
- Static Next.js website: no database to attack, pre-rendered pages, automated tests and HTTPS — the same architecture as this site.
- Search set-up: Google Search Console, Merchant Center product feed, Bing and IndexNow; GA4 key events, with two tracking bugs found and fixed.
- Owner hub on the local PC: live site status, orders from Google Sheets, and a weekly report from GA4, Search Console, Merchant Center, Meta and YouTube that explains each problem and its fix.
- Auto-publishing from one folder and one JSON file to YouTube, Facebook and Instagram via official APIs, with a dry run, duplicate checks, AI-disclosure labels and logs that record a post only when the platform confirms it.
- Backups with a tested restore, secrets kept outside the project, and resumable jobs.
Outcome
- A store, search and social presence, reporting and publishing pipeline operating within ten days.
- Engineering patterns reused for Rehbar: static site, tested build, owner dashboard, approval-gated automation, honest logs.
Evidence
- Static pages
- 530+
- Automated tests
- 23
- Product sizes in shopping feeds
- 1,150
- Pages submitted in sitemap
- 430
- Platforms publishing from one file
- 3
- Software and API cost
- Rs 0
Source: Shopskon case study, 7–16 October 2026 (project logs and analytics).
Limitations
- Figures are from the project’s own logs and analytics on 16 October 2026 and describe an e-commerce operation, not research outcomes.
- The project is recent; long-term commercial results are not yet known.