Hey PH! I'm Christian, one of the developers of Trench.
We're building Trench as a modern open-source analytics infrastructure designed for tracking events, page views, and user identities, powered by ClickHouse and Kafka.
Why did we build this? At our startup, Frigade, we hit a wall with our Postgres-based events table, which became costly and slow to scale as our user base grew. It's a common problem for companies scaling up: the initial events table in a relational database works for a while but quickly turns into a bottleneck when scaling to 1M+ users. Trench aims to solve this pain point with a scalable, production-ready tracking system that keeps costs down and performance up.
Some of the main features include:
1. 💼 Segment compatibility: Fully compatible with the Segment tracking spec (e.g., track(), identify(), group()), so you don’t need to rewrite tracking code.
2. ⚡ Built for scale: Trench is designed to handle thousands of events per second on a single node, ensuring fast and efficient event tracking.
3. 🔍 Real-time queries: Run analytics with read-after-write guarantees in real-time, which gives a huge edge for live insights.
4. 🚀 Simple deployment: Just one Docker image that packages everything you need—no need to manage Kafka, ClickHouse, or node setup.
5. 🔌 Seamless integration: Plug into any cloud-hosted ClickHouse or Kafka provider like ClickHouse Cloud or Confluent.
Use Cases:
- Real-Time Monitoring: Set up alerts for critical metrics like error rates or usage spikes and get notified instantly.
- Event Replay & Debugging: Capture every interaction for easy event replay and debugging.
- A/B Testing: Use real-time segmentation and event capture to tailor experiences on the fly for different user groups.
- SaaS Product Analytics: Integrate Trench with your SaaS product to power audit logs or user tracking features.
- Custom RAG (Retrieval-Augmented Generation) Models: Use event data in real-time to power search or personalized responses with SQL queries, which can enhance your app's AI capabilities.
Trench is open source (MIT licensed), and we're exploring future features like Elastic Search integration, direct data exports (Redshift, S3), and an admin UI for managing queries and webhooks.
Would love to hear your thoughts on how you’ve tackled the scaling challenges of event tracking or if you have any feedback on Trench!
