Hab jetzt keine Zeit selbst einen längeren Text zu verfassen...aber hier ist es eigentlich ganz gut erklärt:
Former PLTR Engineer DD Part #2: Usability, Deployabilty, Scalability, & Submersibility + My Lockup Plan
https://www.reddit.com/r/wallstreetbets/comments/...part_2_usability/
#5) Does it Matter that Palantir Needs FDEs at Deployments?
One of the favorite arguments of the Palantir bears is that Palantir is an unscalable, overvalued services company that should have the same P/E as Accenture, Booz, BAE, etc. The bears say that Palantir only has a handful of clients and can’t scale. In my mind, the quality of the admin tools show on Demo Day and the IBM partnership announcement eviscerated that argument.
When I was at Palantir, it was a bitch to build and manage deployments. From what I saw on Demo Day, Palantir is now easier than Salesforce to deploy. I say that as someone who has written tens of thousands of lines of code for Salesforce (Lightning, Apex, SOQL and Visualforce).
Eventually, Palantir will have the same ecosphere of 3rd party developers that deploy and manage instances of SAP, Salesforce, Microsoft, etc. but that will take years to fully mature that ecosphere. Eventually, Palantir will have robust certifications for various skills and tasks to manage Palantir deployments.
But in the meantime, Palantir's Forward Deployed Engineers (FDEs) will continue to directly manage deployments. Given how much progress Palantir has made on the tools to improve the efficiency of the FDEs, I no longer worry about the scalability of that team.
Say that three utility companies all buy Palantir – PG&E, ConEd, and Entergy. Palantir invests 4 years worth of FDE time in PG&E, 1 year in ConEd, and none in Entergy.
PG&E improves decision making by 10% and does a massive enterprise deal with Palantir to deploy to every part of the organization.
ConEd improves decision making by 2% and does a moderate renewal for a few specific teams.
Entergy bungles their deployment and actually makes worse decision as a result. They do not renew.
Assuming ~$500k/year all in is the cost for a good FDE at Palantir, we're talking about small up front investments that return bigly for PLTR in the long run as those deployments are extended like the BP deal.
To put it another way, when you think about Palantir deployments in months or years, the FDE overhead is spooky. But when you think about those deployments converting into decade long operating system for XXXXXX deals, the FDE overhead is not significant.
Also, to be honest, there are very few people with TS/SCI security clearances that even know what React is, much less how to write code against it, so in some of the more sensitive use cases for Palantir, I think that FDEs will be the way for a long time.
|
1 |
... |
114 |
115 |
|
117 |
118 |
...
| 255
Thread abonnieren
