Blog
Product notes, guides and engineering write-ups.
- ai
- search
Finding a document by what it means, not the words it contains
pgvector for semantic search, fused with full-text search, plus a link graph between documents: what that changes for an agent that has to find the right information.
Read- self-hosting
- postgresql
What a pg_dump changes in a contract negotiation
TblFlow's self-hosted deployment is paid, on quote. What's free, on every tier: real portability of your data via a standard pg_dump, with no conversion.
Read- views
- product
Why switch views instead of duplicating the table
Grid, Kanban, Gallery, Calendar, Gantt, Form: TblFlow's six views read the same table without copying a single row. What that actually changes for a team.
Read- ai
- agents
- automation
When there is no API: the agent takes the browser
Half the tools a small company runs on expose no usable API. Here is how our agents drive a browser instead, and the guardrails without which that approach is a bad idea.
Read- postgresql
- architecture
Why TblFlow writes to real Postgres, and what that actually costs
Most no-code databases store your data in a format of their own. TblFlow creates one real PostgreSQL table per table. Here are the concrete consequences, including the unpleasant ones.
Read- ai
- agents
An agent that acts is not a chatbot with write access
Giving a model permission to change your data does not make it a reliable agent. What else it takes: persistent memory, a proposal cycle, and a human in the loop at the right points.
Read
No articles match this filter.