Inspiration
Inbox overload eats time and attention. We kept hearing the same two pain points from students and teams: important mail gets buried under newsletters and receipts, and follow-ups slip through the cracks. Email365 was born to do both jobs at once—automatic triage and reliable reminders—in a minimalist UI that stays out of the way.
What it does
Email365 connects to your provider and automatically classifies messages into Priority, Personal Finance, Admin, Newsletter, and Spam. A lightweight rule engine lets you add conditions (sender/domain, subject/body keywords, has: attachment, size, date window) and actions (label, move, mark read, star, snooze). Drag-an-email to a category and we prompt: “Always do this?” to create a rule in one click. Reminders are created from detected intents (“invoice due Friday”, “reply by 3pm”) or from quick actions. The dashboard includes powerful search chips, bulk actions, and a two-pane reader. There’s a demo mode, dark by default, keyboard shortcuts, and multilingual copy.
How we built it
We use n8n for our backbones with is the backend, and using caffeine AI to built UI to intergrate with n8n backend, including the gemini API for accessing and resolving the data## Challenges we ran into
Accomplishments that we're proud of
We shipped an end-to-end flow: connect → ingest → classify → remind → act. The rule builder feels approachable, and the “Always do this?” nudge makes learning effortless. The scrollable landing explains use cases clearly, and the dark-first design meets AA contrast. Internal tests on synthetic/public samples show strong early precision, and latency stays snappy with caching and streaming renders. Most of all, users say they feel calmer.
What we learned
Our best experience we gained is a real hand-on project that I have never done it before, also it taught me how to use AI to support our project. Hacktret also a place for me to meet new people, make connections
What's next for Email365
Calendar sync for reminder deadlines, richer provider support, mobile-first tweaks, and a privacy dashboard. We’ll add suggested rules based on confirmed actions, a “follow-up if no reply” automation, lightweight analytics (time saved, response times), and team features (shared rules/labels).
Log in or sign up for Devpost to join the conversation.