Running real-time analytics in Postgres

Running real-time analytics in Postgres is a controversial topic, with a debate between running real-time analytics workloads in combination with transactional workloads or using separate analytical databases. The goal is to achieve better performance by running aggregate queries in a regular OLTP database and potentially eliminating the need for analytical databases like Vertica or Snowflake. However, there are challenges with this approach, as analytical databases are optimized for processing large data volumes and may not perform well with user-facing workloads. Some solutions discussed include the potential for transactional systems to add analytical capabilities and vice versa, as well as the use of hybrid systems that have two separate databases. Additionally, the importance of real-time analytics is highlighted, emphasizing the need for systems that can provide access to data at all times.