The work behind the service
Running a restaurant involves more than orders: shift planning, stock information, timed tasks and helping new staff find the right instructions. This project brought those recurring activities into connected workflows.
What was built
Python connected Trello, Google Sheets and WhatsApp messaging through Twilio. The system supported daily and weekly tasks, shift planning, inventory information and an internal RAG assistant using prepared operating instructions.
Knowledge at the point of work
Staff could ask for guidance on recurring tasks and pizza preparation, including during onboarding. The assistant drew on instructions prepared for the business, rather than treating a general model answer as the operating procedure.
What this demonstrates
A practical way to connect automation, team knowledge and day-to-day business needs. The project informs our approach to choosing a useful first workflow, defining what the assistant can do and preparing a handover people can use.
Project context
This was built for the founder’s own business, not an external SundAI client engagement. This overview is based on the founder’s account. The visual is a simplified illustration; no customer records, measured savings or independently audited performance figures are presented.