- Created tables for AI benchmark runs and results to facilitate model evaluation.
- Added expected answers column to benchmark results.
- Introduced tables for internet connection change cases and vTiger archive management, including records, relations, and checkpoints.
- Implemented triggers to enforce append-only behavior for vTiger archive records and files.
- Enhanced solution management with soft delete capabilities for sag_solutions and knowledge_articles.
feat(scripts): add CRM benchmarking script for Ollama models
- Developed a Python script to benchmark CRM models exposed through Ollama, including various test cases and scoring mechanisms.
test(tests): add comprehensive tests for new features
- Implemented tests for internet change case service, vTiger archive functionality, and sag solution knowledge management.
- Ensured coverage for edge cases and error handling in the new features.
1cfe5aee76