OLTP (Online Transaction Processing) describes the workload of operational databases behind live applications: many small, fast, concurrent reads and writes (place an order, update a profile). It contrasts with OLAP (Online Analytical Processing) — few large aggregate queries over lots of data.
Worked example: the checkout writing a new order row is OLTP (row-oriented, indexed, low-latency); the finance team summing revenue across all orders for the year is OLAP (columnar, scan-heavy), usually run against a separate warehouse. Gotcha: the two workloads fight each other on one system — a heavy analytical query can lock up or slow the transactional path — which is why they are typically split, with data piped from OLTP to OLAP via ETL or CDC.