AI · Productivity · Full stack

Live portfolio demo

DayPilot AI

Plan the day. Understand the trade-offs. Stay in control.

DayPilot AI planner showing a weather-aware schedule, task controls, and workload guidance

Overview

What the project is designed to do

A local-first, weather-aware workday planner with flexible tasks, date isolation, conflict and free-window analysis, and review-only planning proposals.

Tasks support fifteen-minute timing, priority, category, notes, status, and weather sensitivity while versioned local storage provides migration, import, export, and undo.

Open-Meteo context and optional AI wording feed a guarded proposal flow that must be previewed and revalidated before the user applies it.

Problem

What needed to change

Simple hourly schedulers do not explain overlaps, workload, weather trade-offs, or what an automated planning suggestion would change.

Solution

How the release responds

Model the workday as local user-owned data, calculate conflicts and free windows deterministically, and keep every suggested schedule change behind explicit review.

Key features

What visitors can use

  • Flexible fifteen-minute tasks
  • Conflict, workload, and free-window analysis
  • Opt-in weather and air-quality context
  • Reviewable planning proposals
  • Versioned local data with import, export, and undo

Implementation

How it was built

  1. 01

    Next.js App Router and TypeScript separate planner-domain logic from presentation.

  2. 02

    Server route handlers normalize Open-Meteo data without exposing credentials.

  3. 03

    A deterministic assistant remains available when the optional AI SDK path is disabled.

Architecture

Delivery shape

A Next.js application keeps planner data in the browser and uses narrow route handlers for normalized weather context and optional proposal wording.

Product boundary

What this release does not claim

  • No accounts, database, analytics, ads, background location, or assistant history
  • The assistant cannot autonomously change a calendar and weather context is not an official warning

Continue exploring

Related projects