You might have seen our launch (https://news.ycombinator.com/item?id=48195021) earlier. For those that haven’t, we’re building a simpler, cleaner version of platforms like Datadog or Sentry, that triages and fixes the bugs it finds.

We started to work on this project when we understood that we weren’t opening our traditional monitoring tools anymore and relied on their MCPs instead. We also started to feel like setting up dashboards and fine-tuning alerts in the UI was becoming a drag.

Here’s how Superlog is different from other tools:

Every issue is triaged: we use heuristics and LLMs to group error logs and issues into incidents to reduce noise and context switches for on-call engineers

Zero clicks: all of the platform features are available via our MCP server, so you don’t have to spend time configuring your dashboards.

We know that this will be a very useful tool for orgs of any scale, and we want to build it together with the community.

Our ask: we want your suggestions, feature requests, and your feedback.

We’re building on the shoulders of two open source projects that I want to thank.

First of them is ClickHouse, which is an extremely versatile OLAP database. Our friends are trying out CH in different projects, from YC startups to industry leaders, and we have heard very positive feedback.

Another shout out goes to Better Auth. Before BA I often had to choose between a batteries-included experience like Clerk (at the cost of lock-in and dependency), and a free and open-source product like NextAuth/Auth.js (at the cost of some rough edges), or rolling our own.

For Superlog, we could just use BetterAuth, get impersonation and orgs out of the box, without a dependency on a third-party service that would make self-hosting our code more difficult.

We’ll be in the thread for the rest of the day to answer any questions and take in your feedback. Ask us anything!