What's Changed
Features
- feat: add new widget creator page by @Steffen911 in #6322
- feat(dashboards): add beta preview for customizable dashboards by @Steffen911 in #6376
- feat: add metadata field to scores by @davidmc971 in #6046
- feat(run-compare): add detail navigation by @marliessophie in #6398
- feat(evals): Extend job configuration filter options; Evaluators can run based on environments by @davidmc971 in #6039
Fixes
- fix: fix model table provided_model_name by @maxdeichmann in #6356
- fix: update dashboard column accesses after query builder updates by @Steffen911 in #6363
- fix: flaky playwright by @maxdeichmann in #6365
- fix(playground): handle undefined content with toolCalls by @hassiebp in #6364
- fix(scores/trpc): byid on deleted score resulted in 5xx error instead of 404 by @davidmc971 in #6384
- fix: propagate userId correctly to downstream tables by @Steffen911 in #6389
- fix(prompt-ui): internal type on review chat prompt by @hassiebp in #6391
- fix: update column mapping for categorical score charts by @Steffen911 in #6399
- fix: access all replicas to create dataset score metrics by @Steffen911 in #6407
- fix(dashboards): support all dashboard filters across views by @Steffen911 in #6413
Performance
- perf: Increase batch export timeouts by @maxdeichmann in #6375
- perf: retry failed S3 deletions by @maxdeichmann in #6378
- perf(batch-exports): make page size configurable and set to 500 by @hassiebp in #6385
Build
- build: update end to end tests after dashboard change by @Steffen911 in #6406
Styles
- style(compare-runs): remove duplicate run column selector by @marliessophie in #6403
Chores
- chore(deps): bump next from 14.2.25 to 14.2.26 by @dependabot in #6316
- chore: upgrade tiktoken by @maxdeichmann in #6357
- chore: upgrade playwright tests by @maxdeichmann in #6359
- chore: add traces domain layer by @maxdeichmann in #6253
- chore: refactor observations domain by @maxdeichmann in #6353
- chore: create scores domain by @maxdeichmann in #6360
- chore: clarify titles in dashboard latency charts by @Steffen911 in #6370
- chore: Increase delete queue processing observability by @maxdeichmann in #6371
- chore: increase batch export retries and improve observability by @maxdeichmann in #6374
- chore: improve project deletion logs by @maxdeichmann in #6373
- chore: Increase project deletion retries by @maxdeichmann in #6372
- chore: remove beta flag from blob storage integration by @Steffen911 in #6397
- chore: add dummy route that accepts all otel metrics calls by @Steffen911 in #6408
- chore: disable error log checks in playwright tests by @maxdeichmann in #6414
Full Changelog: v3.48.1...v3.49.0