We unified our AI stack into one Postgres instance - and cut our infra and backend code by 90%!
Most teams use 3+ databases already:
- A relational / document store like mongo or postgres,
- A vector DB like Chroma or Pinecone for embeddings,
- Redis for caching, and
- Usually an analytics warehouse, like Athena
You of course have to write and maintain all the glue code & data transformations for each of them. But more importantly, you have to take care of the schema evolution which can be a nightmare when dealing with multiple databases!
Instead we run everything on TigerData (creators of TimescaleDB)’s PostgreSQL platform (TimescaleDB + pgVector + pgAI).
Most engineers on my team are feeling let down with the AI hype. Vibe coding makes some mistakes and does a good job of hiding the things it gets wrong.
They spend more time spotting and fixing bugs and basically have been feeling frustrated.
It's also annoying for the team in general. Projects that would otherwise take a couple of days have sometimes taken over 2 weeks, and it is hard to predict how long something will take. That adds a lot of pressure for everyone.
I think people don't realize how much models have to extrapolate still, which causes hallucinations. We are still not great at giving all the context in our brain to LLMs.
There's still a lot of tooling to be built before it can start completely replacing anyone.
why do you think sonnet is better than opus on this?