Track new features, improvements, and fixes to the Mardii platform. We ship weekly and document everything.
We ship new features and improvements every week. Here's what's new.
Dedicated pages for every monitored AI provider with risk scores, change timelines, and alert subscriptions.
Provider directory page with sortable risk scores and breaking change indicators
Per-provider profiles with change history, weekly frequency charts, and risk breakdowns
Follow any provider and sign up for email alerts on score changes
Centralized provider data layer powering all intelligence views
Redesigned calculator with support for complex usage patterns, batch pricing, and multi-region deployments. Now includes confidence intervals for all projections.
Added support for batch API pricing calculations
Multi-region cost analysis for global deployments
Confidence intervals on all financial projections
Integration with 12+ cost management platforms
Enhanced risk scoring algorithm with improved model lifecycle risk assessment and better handling of policy drift detection.
Model deprecation notices now weighted by migration complexity
Policy drift detection improved with clause-level analysis
Historical data window extended to 18 months for stable vendors
Fixed webhook delivery issues and improved retry logic for failed notifications.
Webhook delivery success rate improved from 97.2% to 99.8%
Added exponential backoff for failed webhook deliveries
Webhook endpoint health monitoring and automatic retries
Better error messages in webhook delivery logs
Added monitoring support for Google's new Gemini 2.0 models and updated pricing detection for the new tier structure.
Gemini 2.0 Flash and Pro models added to monitoring
Updated rate limit detection for new Google API structure
Historical data backfilled for Gemini 1.5 deprecation timeline
Increased API rate limits for Business tier customers and added new Enterprise tier options.
Business tier: 1,000 → 2,500 requests/hour
Enterprise tier: Custom limits up to 50K requests/hour
Rate limit headers now include reset time and remaining quota
Fixed false positives in Anthropic's ToS monitoring and improved detection accuracy for OAuth policy changes.
Reduced false positive rate for Anthropic ToS changes by 85%
Added specific detection rules for OAuth and API key policies
Improved change classification for developer tool restrictions
Major platform update with redesigned dashboard, new financial tools, and enhanced API access.
Completely redesigned dashboard with improved UX
New Financial Impact Calculator with scenario modeling
ADES Portfolio Risk scoring for concentration analysis
REST API v1 with webhook support
Enhanced AVRS methodology with 5-factor scoring
Real-time change feed with <4 hour detection
Want to stay updated on new releases? Sign up for email notifications