Unified Observability in Obics

The premise and strength of Obics is the ability to do advance analytics with all telemetry signals. That includes logs, metrics, traces, and whatever custom events you might want to use. To deliver on this promise, Obics uses a single distributed ClickHouse database for all telemetry. That means all your services and clients send logs, metrics, and traces to the same place. This allows to correlate between different systems and telemetry types. For example, you can find out that error logs are related to high CPU metrics. Or that traces with a p99 latency are always during high memory utilization and always contain a specific log message like "[WARN] Memory usage is high".

You might think that this is great in theory, but writing long SQL queries to utilize these capabilities is not an option. That's why Obics includes a custom-tailored AI assistant that will write those queries for you, and even gather insights from the results. You can ask "Do I have any memory problems in my backend services?" or "See if there's correlation between high CPU times to high latency in traces with request url /transfer-payment. If so, which service have the relevant CPU spikes?". Obics AI knows the structure of your logs, metrics, and traces. It's fine-tuned on your specific data and knows how to construct the right queries.

Want to see it to believe it? Try the sandbox