Mnemiq
Open-source text-to-SQL you can tune and measure on your own database
Emerged finds new products worth a look, daily — what’s new today
It predicts on messy spreadsheets and relational tables better than a tuned XGBoost most of the time, with little of the feature engineering.
TabPFN-3.5 from Prior Labs is a foundation model for prediction on structured data. It ranks first on TabArena and BeyondArena and is aimed at the tables that break classic ML: text-heavy columns, high-cardinality categories, temporal and grouped splits, and relational data. It comes as Plus, Thinking (most accurate) and Fast (up to 6x faster, alpha) variants through the Prior Labs API, SAP AI Core and AWS SageMaker, while the base and Fast checkpoints are open through the tabpfn Python package and Hugging Face.
More like this
TablissNG
A revived Tabliss new-tab page with more widgets, backgrounds and search options
Warp
Run trillion-parameter models on a laptop by streaming weights from NVMe
TabMail
AI triage, summaries and drafts for the inbox you already have, in open clients
Rustable
A free desktop SQL and Redis client with no cap on open tabs
xarray-sql
Run SQL over Xarray datasets with DuckDB, DataFusion or Polars