Okay, listen, I love NYC but many people here love thinking they’re pushing out Berlin-levels of fashion when instead it’s just a horrendously trash-ass fit.
@pdrmnvd Just made a new one for a board meeting, I call it "the cobbler's lament". It's the first of its kind and I won't ever recreate it for another presentation as long as I live.
We should be stealing a lot more from data engineering when we build AI systems.
DAGs - does this really need to be one giant agent loop?
Lineage - which sources, artifacts, prompts, and model versions produced this?
Change detection - if a source changes, which generated artifacts are now outdated?
Incremental processing - what can we keep instead of rerunning the whole agent?
ETL - what context do we pull in, what does the model do with it, and where do we save the result?
Data engineers have been solving these problems for decades.
I love my Garmin Fenix watches, I’ve owned several, their hardware is wonderful.
They make the biggest heaps of steaming shit software like you could never believe. Bluetooth and WiFi issues like no tomorrow. Having to reset and re-pair the devices all the time.
“Postgres doesn’t scale well”
Where is your love of the game. The unbridled optimism to figure it out, learn how it works, and hyperoptimize a system. Get pgstats and pgbouncer rolling, and start digging in the query plan.
Only then you may pivot to DuckDB and DynamoDB