Hacker News new | ask | show | jobs
by swordsmith 613 days ago
I use Foundry for work. It makes data ingestion, cleaning, quality check and automation easy. After all the data is ingested, running analysis/RAG on them become extremely easy.

Basically, it's end-to-end data engineering and analytics. And the more a company uses/invests into the platform, the more benefit and locked-in they are.

3 comments

"End-to-end data engineering and analytics" is quite a bold claim from a single service provider.

Here is the link for anyone interested: https://www.palantir.com/platforms/foundry/ and a YouTube explainer: https://www.youtube.com/watch?v=ZGGRCTTjLfQ

Given you've used it, just how self-service is it? To me this seems like such a large claim that - if it's doable - I'm surprised there are not more competitors in the "vertically integrated data providers" space.

> Given you've used it, just how self-service is it? To me this seems like such a large claim that - if it's doable - I'm surprised there are not more competitors in the "vertically integrated data providers" space.

It is both very self service and not very self service. That's why they employ the FDE model from the article, to actually ingrain it into the client company to the point that it becomes self service.

It's extremely hard to build such a product from scratch and have it actually be good, that's why there's no competitors. Especially providing the finely grained security controls that the article talks about, and have the platform be secure. There's a reason their security team wins the biggest CTFs half the time.

It is completely self service by now. I have my own stack for testing purposes. Of course if you want to deploy this to an enterprise things will differ, but that is the same for Snowflake, Databricks etc.
It certainly sounds like they've created an excellent product both for its value to the customer as well as value to their shareholders.

That's what companies should all be built and optimized to do. That's what it's about.

RAG?
Retrieval Augmented Generation.

https://en.wikipedia.org/wiki/Retrieval-augmented_generation

Basically, using your actual data/documents to supplement a general purpose LLM and generate better answers for your specific use case.