1,664 karma · joined October 16, 2011
@sbourke
Or people do some fine tuning of a model or their own and store it someplace (maybe es). In which case you could still be using es to cover filters and range parts of the queries and then rerank that result set.
Not covered the ops and scalability benefits you get with elastic / solr / open search which is also very important if you’ve non trivial usage
It involves stating the type of position someone has, the tier company someone works at, which somehow demonstrates the persons value = (position * tier). Then because this is the norm people this is how people befriend each other, or how they attract love interests.
I'd just chalk it up to human nature.
What I value about these lls is that they are essentially a condensed version of the internet (despite being stupidily large for normal hardware currently)
Usually if I’m building a recommendation or search algorithm I have to use the data from the company I’m working with. This makes it possible for me to encode the entire internet into a model that might be running on a product that only has 100 users.
Doesn’t look like the individual has access to budget or autonomy to make significant purchases (I.e no good path for backhanders from vendors or consultancies)
headline feels sensationalist
But what I often miss from posts like this is focusing on the value the group of individuals is meant to create (maybe it’s implied).
There will be no explanation of the actual algorithm.
Inverted indices with fuzzy and boolean matching are pretty boring, but they're still pretty trivial to stand up and do not require anything beyond the actual corpora to be built.
The way I would read that is they wanted to hire someone in at the earlier stages of their career rather than more experienced. It can be important to have a balance of seniority in a team.
Fast forward 20+ years later, that type of work is essentially the SAAS apps you see popping up. It's not a very capital intensive business to sustain so with a few thousand customers you could have a pretty healthy company and continue product development / grow your customer base.
Why do people go for these new SAAS apps ? I've seen a few reasons
1) Pricing is somehow better 2) Integration and or support is more personal so the customer (usually a business stakeholder) prefers it. 3) Some specific feature exists that is important to the customer i.e ease of deletion for GDPR requests 4) CTO / CIO / Techie prefers it.
Why I've seen people not go with the open source alternatives is the same oul reason. Focus. If you've less than 30 people at your org, unless one of those tools has a specific team covering that area - it will just split focus and end up in a suboptimal setup and create pain points in terms of support.
Examples that come to mind, we used open source snowplow analytics because we had a data platform team that supported it and ran it full time. There was no cost with regards to the software, but certainly was with head-count to maintain, and evolve.
We used external newsletter services to inform customers about updates etc. in the product. We went with an external paid service because this didn't tie directly into a specific team and for 100 - 500 euro a month we got what probably would have cost 0.3 - 1.0 full time head count to keep a minimal service up and running (Which meant -0.3 to -1.0 headcount working on important things for the company).