Question 1

Which Front objects accept writes from Databricks?

Accepted Answer

Accounts and contacts. Everything else is source only, including conversations, channels, contact lists, inboxes, tags, teammates, teams and events, so those can be read into the lakehouse but not written to. Accounts and contacts are the practical targets anyway, since they hold the context somebody needs the moment a message lands.

Question 2

How do I define what gets sent?

Accepted Answer

Either by selecting a catalog, schema and table, or by writing custom SQL which Polytomic wraps as a subquery and runs against your SQL warehouse. For a destination like Front the query is usually small and specific, something like current customers with their plan and renewal date, so custom SQL tends to fit better than pointing at a wide table and mapping a handful of columns out of it.

Question 3

What access does the connection need?

Accepted Answer

Read access only, since staging storage is a destination concern and this direction never touches it. Point the connection at your SQL endpoint with a personal access token or an OAuth service principal, and give it permission to read the catalog and schema the model uses. Nothing about this sync needs write permission in Databricks, which makes it a reasonable candidate for a narrowly scoped identity.

Question 4

Should every customer in the model end up in Front?

Accepted Answer

Almost certainly not. If the model covers everyone you have ever recorded, update or create imports all of them into a shared inbox where they are noise rather than context. Update only is normally right here: it enriches the contacts Front already knows about because somebody actually wrote in, and leaves the rest of the lakehouse where it is. The inbox stays about people who are talking to you.

Question 5

What does an agent gain from this?

Accepted Answer

The ability to tell conversations apart before reading them. A shared inbox shows an address and a message, so everything looks equally urgent until somebody investigates. Reverse ETL attaches plan, spend, usage and renewal date to the contact, so the queue can be ordered by what actually matters and a reply written knowing who is asking, which is a different thing from simply replying quickly.

Sync Databricks to Front

Sync Databricks to Front with Polytomic

Contacts and accounts

Table or query

Matching prevents duplicates

Or sync Front to Databricks

Trusted by businesses worldwide

Step-by-step integration guide

Connect Databricks in Polytomic

Connect Front in Polytomic

Create a sync between Databricks and Front

Map fields and enable the sync

Frequently asked questions

The sync and ETL platform that does it all

Powerful data sync

Advanced analytics

Enterprise security

Easy setup

Ready to connect Databricks with Front?