5 ms·

To expand on your point on Palantir just being a database, it's a good lesson to startup founders: solve the problem your user had. The biggest problem that Pa

by tyre 10y ago

To expand on your point on Palantir just being a database, it's a good lesson to startup founders: solve the problem your user had. The biggest problem that Palantir solves is: we (CIA, FBI, whatever) have to pull together data from 500 software systems that don't share a common taxonomy, hold petabytes of data in totally different ways, live in agencies across the world, and have all sorts of nuanced permissions models (e.g. "Five Eyes" or "this data was shared to New Zealand by a country that doesn't share data directly with the US, but New Zealand does share data to the US. I'm a US agent looking at NZ data, do I see this?"). They basically clean and centralize information. It's super unsexy (probably my least favorite thing to do as an engineer.) But that is the problem that these organizations have.

piker 10y ago
Very good points. Any insight as to how Palantir learned so much about its customers' problems? Does Palantir employ a good number of its customers' former analysts, etc.? Just curious how this is done so well in practice. [edit: clarity]
- astronautjones 10y ago
  It was funded by In-Q-Tel (the CIA's VC firm) from the start
- pcsanwald 10y ago
  They behaved basically like a consulting shop. send tons of "forward deployed engineers" onsite.
tluyben2 10y ago
> To expand on your point on Palantir just being a database, it's a good lesson to startup founders: solve the problem your user had. Exactly. You can find this in almost any 'stodgy' business around. If you see very little computers around or only computers with glowing Word screens, you can be sure, if you find some business person to talk with, he or she can tell you what really would boost their business. And then you suddenly have a business if the competitors also do not have it (usually they don't). Gov is a good one as they are usually slow and not very creative (generally! don't want to step on toes but what I saw so far in the world shows me that this is generally true).

by tyre 10y ago

piker 10y ago

Very good points. Any insight as to how Palantir learned so much about its customers' problems? Does Palantir employ a good number of its customers' former analysts, etc.? Just curious how this is done so well in practice. [edit: clarity]

astronautjones 10y ago
It was funded by In-Q-Tel (the CIA's VC firm) from the start
pcsanwald 10y ago
They behaved basically like a consulting shop. send tons of "forward deployed engineers" onsite.

tluyben2 10y ago

> To expand on your point on Palantir just being a database, it's a good lesson to startup founders: solve the problem your user had. Exactly. You can find this in almost any 'stodgy' business around. If you see very little computers around or only computers with glowing Word screens, you can be sure, if you find some business person to talk with, he or she can tell you what really would boost their business. And then you suddenly have a business if the competitors also do not have it (usually they don't). Gov is a good one as they are usually slow and not very creative (generally! don't want to step on toes but what I saw so far in the world shows me that this is generally true).