Processing Delay in Kinesis -> Lambda -> DynamoDB

This is my first time asking a question.

Putting data to Kinesis in AWS Lambda is developing a write to DynamoDB.

The DynamoDB has enough write capacity units, but
Processing speed cannot keep up with data put to Kinesis
The wait time for processing to reflect is getting longer.

Kinesis data is sequential, so it may be obvious, but
There seems to be only one Lambda running at the same time
I think I can solve this problem if the message queue and the worker are one-to-n.
I can't find a way to do it without increasing the Kinesis shard in Kinesis->Lambda...
In the first place, I would like some advice such as using Kinesis incorrectly.
Thank you for your cooperation.

[Operation information]
Kinesis shard count: 1
Put data to Kinesis: 10 to 50 data per second