The solution back then makes sense. It was a couple of years too early for CouchDB, Cassandra or MySQL Cluster (https://en.wikipedia.org/wiki/MySQL_Cluster) which are more suited for write-heavy applications and clustering across servers.
Later spinning harddrives were replaced by SSD, then NVME. More open source NoSQL and columnular storage solution. Cloud services started offering hosted databases with unlimited scale. 500GB would all be 'hot' ready to be queried in realtime.
Today I'd see three options
* multiple cloud servers which receive the data and put it into a managed cloud database, like Google BigQuery. They'll handle all scale, including region replication, backups. You might overpay but likely still less than Oracle software licence.
* specialist SaaS for IoT, for example ClickHouse. They can handle 10.000 incoming rows per second. The data store later does defragmentation, storing data by date and other optimizations which make it faster to query recent data, vs older data.
* place it into JSON or CSV files, one per hour, or one day and query with DuckDB.
If a company gives away their product for free (open source) and tries to upsell users to a pro version on their documentation pages. And if AI answers user questions and users no longer have to browse the documentation pages. Then the company will likely make less money.
I don't know if the company needed 4 software developers, they were working on a big new release, but revenue projections were down and the business long-term not sustainable anymore. A brand new company might be much more conservative hiring. A company with bigger teams more exposed.
"Pre-launch (3 days before):
Posted on Twitter: "Launching Nexus on Friday -
AI memory for multi-agent systems"
I can't find such tweet on the author's twitter feed. No mention of a launch on his linkedin feed either.
"Posted on Product Hunt at 12:01am PT"
I can't find a producthunt launch for Nexus.
"Shared to Twitter, Indie Hackers, relevant Discords"
I can't find a single tweet mentioning trynexus.vercel.app
"What Didn't Work [...] $10/mo signaled "cheap tool." When I raised to $25/mo, nobody
complained"
trynexus.vercel.app still lists $10/month.
On the website all links to subpages are empty. Pricing says "coming soon", only payment method is Zelle which would be direct bank transfer (one-off, not a subscription).
Afaik 'Launch HN' is reserved for YCombinator companies. It gives an automatic ranking boost and guaranteed to hit the frontpage. HN is run by YCombinator. Not sure how the website figures out who belongs to YCombinator. I've read there's a username list and the names show up in a different color so users recognize each other.
Based on using the employee badge in office buildings it seems.
"The updated dashboard, which began rolling out in December, allows managers and HR to view how often employees come into an office, how long they stay, and the locations where they work."
Fun fact: During the Mar-a-Lago search, the FBI had to use agents with appropriate high-level security clearances. Some of the documents were so sensitive they were marked Top Secret/SCI. This is normally even more restricted level than just Top Secret.
Given that "SCI" stands for "sensitive compartmented information", I bet they had to find agents currently read into certain "compartments", or read new agents into those "compartments".
My previous statement was imprecise. They need to be eligible but only those assigned to intelligence, counterintelligence, or national security roles are actually granted access to specific SCI compartments.
The article below mentions the use of taint teams, meaning only agents with appropriate clearances, need to know, and role assignments served on these teams. So not every FBI agent present on the operation looked at classified materials.
