01Problem
Repertory theatres plan every month by hand. The coordinator juggles dozens of titles, actors playing in several venues at once, guest performances, finance targets and last-minute cast changes — usually in a spreadsheet, over email and phone. One missed conflict means a cancelled show.
02Role
We own the product end to end: we brought the concept to the client, designed the UX and interface, defined the data model and roadmap, and deliver the software itself using AI-assisted development — product management, design and delivery under one roof.
03Process & product
- A generator that learns the theatre. The planning engine learns each stage's profile from past sold seasons — density, weekday rhythm, curtain times, seasonality of titles — and proposes month drafts in four variants aligned with finance goals.
- Hard conflict guard. Cast and understudy assignments are checked against availability and cross-stage clashes at every save — double-booking is structurally impossible, not just discouraged.
- The whole workflow, not just the calendar. Actor availability collection, messaging with read receipts, automated weekly/monthly notifications (e-mail and SMS), finance forecasting per repertoire variant, Excel import of legacy schedules, and GDPR compliance.
- Multi-tenant by design. Each theatre organisation runs isolated on the same instance, with per-stage catalogues, casts and learned profiles.
Planning — generation conditions, six-month status board, repertoire variants per finance goal.
Repertoire — an approved month: 20 titles, 55 performances, cast per show.
Finance — forecast per repertoire variant, built on real price data.
04Outcome
The planner is in pilot use by Warsaw theatres. The coordinator reviews and edits AI-drafted months instead of building them from scratch; actors confirm availability and receive schedules automatically.