Skip to content
Notifications
Clear all

Thoughts on the new 'Claw Lite' tier? Might make TCO work for pilot projects.

1 Posts
1 Users
0 Reactions
1 Views
(@data_pipeline_newbie_42_v2)
Estimable Member
Joined: 2 months ago
Posts: 106
Topic starter   [#9543]

Hey everyone, been lurking here for a while and learning a ton from your TCO breakdowns. 😅

I saw the announcement for Claw's new 'Lite' tier and got really curious. My team has been wanting to pilot a new analytics pipeline (think: Airflow for orchestration, dbt for transforms, pushing to BigQuery) but the full Claw platform was always a non-starter cost-wise for just a pilot.

The Lite tier seems to cap compute hours and has fewer concurrent run slots. My immediate thought was: this could finally make the TCO work for a 3-6 month proof-of-concept.

* We could test the integrated lineage features they're always talking about.
* See if their managed Airflow is easier than our current hacked-together solution.
* Prove value before fighting for a bigger budget.

But I'm trying to be realistic. Has anyone run the numbers yet on a similar pilot scenario? I'm worried about:

* The "per-seat" cost for the Lite tier – if we need 5 analysts to just *look* at the lineage, does that blow the model up?
* Hidden egress costs getting data out to our warehouse.
* Whether the concurrency limits mean our pilot feels artificially slow and gets a bad rep.

I attached a screenshot of my super rough first-pass spreadsheet (column H is where I'm stuck on those per-user costs). Maybe someone with more experience can point out what I'm missing?

The dream is to have a pilot that's actually representative without needing a full-scale budget. Could Lite tier be the answer, or are we better off with a frankenstein stack of open source for the pilot phase?


null


   
Quote