A job marketplace and professional social network for the UK construction industry — with an AI agent living in WhatsApp.
Job marketplace · Social platform · AI agent2026- Role
- Sole backend architect & developer
- Company
- Fixed Construction
- Year
- 2026
- Stack
- PythonDjangoREST APIsWhatsApp Business APICron automationGCP
- ~35%
- increase in daily active engagement
- 8–10h
- manual work saved per week
The problem
Fixed Construction is building a dedicated home for the UK construction industry: a job marketplace and professional network connecting tradespeople with employers and site work. It needed a backend that could carry a job board, a live social feed, worker profiles, and trade-specific CVs — plus distribution channels that didn't depend on someone manually posting jobs all day.
What I built
- 01
Architected the platform's Python/Django REST APIs end to end — job board, live feed, worker profiles, and trade-specific CVs — with authentication and role-based access control.
- 02
Built the social module: post creation and a friend-request-style connection system between users.
- 03
Engineered a cron-driven social-media automation pipeline that picks the day's highest-paid job and auto-publishes it to Facebook, Instagram, and LinkedIn.
- 04
Built a WhatsApp AI agent that lets users post and remove job vacancies and check application status conversationally — no dashboard required.
- 05
Built Telegram and WhatsApp scrapers that capture external job and lead data into the platform.
- 06
Modernized legacy services and streamlined multi-environment deployments on Google Cloud Platform.
The outcome
The automation pipeline saves roughly 8–10 hours of manual posting work every week, and the WhatsApp agent lifted daily active engagement by around 35%. The platform now runs on clean, multi-environment GCP deployments.