Question 1

Can Polytomic sync from DynamoDB incrementally instead of scanning the whole table?

Accepted Answer

Yes, by enabling DynamoDB Streams on the table. Polytomic then reads changes from the stream, which covers inserts, updates, and deletes, so each ETL run moves only what changed. Without streams, reads are full table scans on every execution. There is no watermark or tracking field mode for DynamoDB the way there is for a relational database, so streams are the mechanism to use if scan cost or table size is a concern. Targeting a secondary index or using a SQL query in a model can also reduce how much of the table is read.

Question 2

How does Polytomic handle a single table design?

Accepted Answer

A physical DynamoDB table appears as one table in Polytomic. Entity types are not split apart automatically by partition key prefix or by a discriminator attribute, so a table holding users, orders, and events together arrives as one wide table. You have two practical options: target a secondary index that isolates the entity you want, or use a SQL query in a model to select the subset and shape it. Many teams do the separation in the data warehouse after loading, which also works well.

Question 3

What happens to nested maps and lists?

Accepted Answer

Map attributes become object fields and list and set attributes become arrays, and this holds for records arriving through DynamoDB Streams as well as full scans. Nesting is preserved recursively rather than flattened into relational columns, so a deeply structured item does not lose its shape in transit. If you want a flat table at the end, do that shaping in your data warehouse or in a model query, where you have full SQL available.

Question 4

Can I write data back into DynamoDB with reverse ETL?

Accepted Answer

Not currently. DynamoDB is a source only connector, on both hosted and self hosted deployments. You can run ETL out of DynamoDB into your data warehouse, cloud storage, databases, and SaaS tools, but Polytomic does not write into DynamoDB. If writing into DynamoDB matters for your use case, it is worth telling us, since connector scope is normally extended on exactly that kind of request.

Question 5

What if some attributes only appear on a small number of items?

Accepted Answer

Schema discovery works from a sample of items in the table and unions the attributes it finds. An attribute that appears on only a handful of items may fall outside that sample and not be discovered automatically. When that happens you can write in the field manually and set its type, and Polytomic will sync it from then on. This is worth checking on tables with sparse entity specific attributes, which is common in single table designs.

DynamoDB

Perfect pairings

DynamoDB Streams capture

Nested attribute support

Analytics ready output

Example workflows

What you can sync

From DynamoDB

How it works

How to set up

Frequently asked questions

Why teams choose Polytomic

No engineering required

Flexible data modeling

Handles scale automatically

Start syncing your data with DynamoDB in minutes

More integrations beyond DynamoDB