CaiLENDAR

Plans arrive as messages, not calendar entries

A booking confirmation, a school note or a WhatsApp message can contain everything needed for a calendar event, but still leave the reader with repetitive admin. They must find the date, interpret the time, copy the location and decide what belongs in the notes. CaiLENDAR starts with that gap between human planning context and structured calendar data.

A draft the user can trust

The product turns pasted planning text into an editable calendar draft. The intended journey is deliberate: provide the context, review sensitive details, generate a draft, check the event and export it. The calendar the user already relies on remains the destination.

Privacy review is part of the interaction. People can inspect and protect sensitive text before AI processing rather than treating data handling as a separate policy page. A generated event remains a draft to review, especially when dates or time zones need interpretation.

Compatibility is the product

ICS output gives the workflow a concrete, portable result. The product does not need to become another calendar to be useful; it can remove the work of preparing a good entry. That focus keeps the value easy to explain and makes the final event more important than the model doing the extraction.

The current story centres on text input. PDFs, screenshots and broader extraction workflows appear in product planning, and are not presented here as universally shipped capabilities. Likewise, this case study makes no promise that every ambiguous date will be resolved correctly or that every plan can be processed in a fixed time.

What this work demonstrates

CaiLENDAR combines AI-assisted extraction, structured output and careful interaction design around an ordinary repeated task. The next validation work is to make uncertainty easier to understand, test privacy review with real users and harden event export across common booking and household scenarios.