Observability is changing drastically with wide use of Opentelemetry in modern applications. Now it helps to direclty derive predictions using ML algorithms, finding anomalous logs, using LLM observability and many such things.Covering how to simply cut down on downtime and troubleshooting time by implementing the Otel Collector, which will ultimately benefit the business.One of the main issues with observability is cost, which is caused by the massive amount of data that is regularly ingested. To lessen this, people are utilizing data ingestion pipelines in the modern day, which only ingest the necessary data while ignoring unnecessary data.I will cover most of the changes that have happened in recent times and also will share some insights on how AI will be integrated to automate some tasks and help in error prediction in observability.