Tickio Keywords Autopinger

Tickio needed a dashboard where their members can manage their monitored Discord channels and keyword rules in one place. Their old setup was a python based monolith, so one large codebase handled everything. That worked at first, but over time it became harder to ship changes safely, scale specific parts of the system, and keep notification logic consistent across platforms. We rebuilt the project with microservices so Tickio could scale faster and maintain the system with less risk.

The Problem We Had to Solve

The old architecture mixed too many responsibilities in one place. Message monitoring, matching logic, notification delivery and keyword and channel management features. This created bottlenecks: one busy component could affect the whole system, and adding a new notifier platform meant repeating logic in multiple places. Tickio needed a cleaner system where each part had one job and could scale independently.

What We Built

Dashboard and API Foundation

We built a Next.js dashboard for managing channels, keywords, and notification endpoints, backed by a Cloudflare Worker using Hono and tRPC. This gave their memebers a modern interface to control their data.

Discord Autopinger Service

We implemented a dedicated discord-autopinger service that listens to Discord events, evaluates messages against keyword rules, applies cooldown logic to reduce noisy repeats, and prepares delivery jobs. Instead of sending notifications directly, it pushes jobs into queues, which keeps the bot fast and focused on detection.

Notification Worker Services

We split notification delivery into separate workers for Discord DMs, Discord webhooks, Pushover, and Telegram DMs. Each worker consumes only its own queue and handles retries, failures, and platform-specific delivery details. This isolates failures and allows each notifier to scale independently.

Why This Architecture Works Better

Moving from a monolith to microservices gave Tickio practical benefits. Detection, API, and delivery are now decoupled, so each service can be deployed and scaled on its own. Queue-based delivery protects the core autopinger from provider slowness.

Results and Metrics

After the rebuild, Tickio saw clear improvements in both speed and scale. On average, each message is now processed and delivered around 1 second faster than before. The platform now handles around 250,000 notifications per month reliably. It also supports a large and growing set of monitored keywords without slowing down the whole tool.

Outcome

This project was not just a dashboard build. It was a full infrastructure rewrite from a monolithic app to a scalable microservice platform. Tickio now has a modern monitoring stack with a clear control dashboard, a fast Discord autopinger core and independent notification workers that keep the whole system running while it grows.