Skip to content
Can Sevim
Back to Side Projects

DailyPulse

Personal AI Automation Project · Ongoing

Overview

DailyPulse is a personal automation project that aggregates industry news from more than 50 RSS sources — spanning AI, fintech and product management — and turns it into structured daily briefings, delivered automatically each morning.

Why I Built / Explored It

Staying current across AI, fintech and product management meant checking dozens of scattered sources every day. I built DailyPulse to solve that problem for myself: turn a manual, repetitive information habit into an automated workflow, and use it as a hands-on way to learn practical AI and automation tooling rather than just reading about it.

What It Does

The system collects articles from over 50 RSS feeds on a schedule, processes and summarizes them with AI, and delivers a structured morning briefing through Telegram — condensing a wide spread of industry news into something I can read in a few minutes.

Product / Implementation Approach

  • Built the core pipeline as a set of n8n workflows: scheduled collection from 50+ RSS sources, processing, AI summarization and delivery.
  • Used OpenAI's API to summarize and structure raw articles into a concise, readable briefing format.
  • Used Google Sheets as part of the workflow for tracking and handling source and content data between automation steps.
  • Delivered the finished briefing through Telegram as the daily reading surface.
  • Self-hosted the automation stack by running n8n in Docker on a low-cost Hetzner VPS, rather than relying on a managed automation platform.

Tools / Technologies

n8n · Docker · Hetzner VPS · OpenAI API · Telegram · Google Sheets

What It Demonstrates / What I Learned

  • Hands-on experience designing multi-step automation workflows, from scheduled data collection through AI processing to delivery.
  • Practical understanding of working with LLM APIs for summarization and content structuring, not just prompting in a chat interface.
  • Experience self-hosting and operating an automation stack (Docker on a VPS) end to end.
  • A concrete example of identifying a personal information problem and building a working automated solution around it.