Skip to content

Release Notes – Version 1.3.1

  • Enhanced Synthetic Data: Introduced enhacements to synthetic generation to improve the diversity and quality of generated outputs.

  • Addressed Minor Bugs: Addressed various minor bugs, including inability to log out of the platform and not just using gold records for accuracy testing.

Release Notes – Version 1.3.0

Dataset Management & Testing

  • Dataset Creation & Management: Comprehensive dataset system for organizing test data with support for both Policy Rules and Agentic guardrails. Create datasets from scratch, import from CSV, or export to CSV for external analysis. Manual record editing with full CRUD operations.

  • Synthetic Data Generation: Parallel bulk generation of synthetic test records (1/10/100 at a time) using configurable LLMs. Advanced candidate scoring with D-score diversity metrics ensures high-quality, diverse training data. Background job processing with real-time progress tracking.

  • Test Execution Framework: Background test job system for running comprehensive accuracy and variance tests against datasets. Real-time progress monitoring with cancellation support. Jobs persist across navigation for long-running test suites.

Quality Assurance & Validation

  • Accuracy Testing: Evaluate guardrail performance against gold-standard datasets with detailed per-rule metrics including F1 score, Precision, and Recall. Overall guardrail scoring with confusion matrix analysis to identify systematic errors.

  • Variance Testing: Statistical variance analysis across multiple test iterations to measure guardrail consistency. Identify non-deterministic behavior patterns and evaluate reliability across different inputs and conditions.

  • Test Results Visualization: Interactive test result displays with sortable tables, per-rule performance breakdowns, and detailed accuracy metrics. Export test results for reporting and quality assurance workflows.

Enhanced User Experience

  • Record Management UI: Intuitive interface for viewing, adding, editing, and deleting dataset records. Context-aware modals that adapt to Policy Rules vs Agentic guardrail types with appropriate field validation.

  • Background Job Monitoring: Real-time status cards for generation and test jobs with progress bars, completion metrics, and cancel/done actions. Jobs continue running during navigation with persistent state.

  • Data Quality Tracking: Automatic labeling of records as Gold (uploaded/manual) or Silver (synthetic) quality for downstream filtering and weighted evaluation strategies.

Release Notes – Version 1.2.0

Authentication & Authorization

  • Authentik Integration: Enterprise-grade authentication system with SSO support and comprehensive user management.
  • Role-Based Access Control: Three distinct roles with granular permissions:
  • Developer: Full access to create, edit, and configure guardrails and applications
  • Reviewer: Can test guardrails in playground and view configurations (read-only for editing)
  • Auditor: Read-only access across the platform for compliance and audit purposes

Configuration Management

  • Version Control for Guardrails: Track changes to guardrail configurations with complete version history and ability to revert to previous versions.
  • Version Control for Applications: Maintain version history for application DAGs and configurations with rollback capabilities.

New Guardrail Types

  • Entailment Guardrail: Advanced faithfulness verification that automatically extracts individual claims from responses and evaluates each claim against provided context using sophisticated entailment scoring. Supports configurable claim granularity (sentence, clause, or fine-grained) and provides detailed per-claim analysis with confidence scoring.

Policy Rules Enhancements

  • Verification Guardrails for Questions: Add secondary verification layer to Policy Rules questions, enabling configurable guardrails to validate LLM-generated answers for improved accuracy and reliability.

Release Notes – Version 1.1.2

Enhanced User Experience

  • Sorting for Guardrails and Applications lists: Quickly organize items in both views.
  • Fullscreen Application Configuration Modal: Normalized to fullscreen for consistency with Guardrail configuration.

Applications

  • Import/Export Applications (from UI): Export an application configuration and import it into another workspace directly from the UI.

Playground

  • Increased Timeout: Extended Playground timeout to better support longer-running experiments.

Platform & Infrastructure

  • Kafka Topic Creation Flow: Adjusted creation process to avoid topic creation bootstrapping issues.
  • Kafka Image Update: Switched to the Apache-maintained image due to Bitnami deprecation.

Release Notes – Version 1.1.1

Enhanced User Experience

  • Speech-to-Text Inputs for Textareas: Add microphone-powered STT to textareas across the UI for faster, hands-free input.

AI-Powered Automation

  • RAG-Based Pipeline for Policy → Rule Extraction: New retrieval-augmented option for converting policy documents into structured rules.

Expanded LLM Support

  • Support for GPT-5: Configure and run guardrails, workflows, and playground experiments with GPT-5.

Reliability & Quality

  • Bug Fixes & UI/Usability Improvements: General stability enhancements and minor UI polish.

Release Notes – Version 1.1.0

AI-Powered Automation

  • Auto-DAG Generation from Policy Documents: Revolutionary feature that uses Large Language Models to automatically analyze policy documents, extract compliance topics, and generate complete DAG structures. Simply paste your compliance document and let AI create comprehensive verification workflows.

  • Intelligent Topic Extraction: Advanced LLM-powered analysis that identifies 3-5 key compliance themes from any policy document, providing the foundation for systematic verification coverage.

Core Features

  • Application-Scoped DAG Workflows: Create sophisticated verification pipelines by orchestrating multiple guardrails with Directed Acyclic Graphs (DAGs). Configure complex workflows that combine parallel execution, sequential dependencies, and custom aggregation logic within application containers.

  • Visual DAG Editor: Intuitive drag-and-drop interface for building DAG workflows with real-time visualization, node property configuration, and seamless switching between visual and JSON editing modes.

  • Interactive Playground: Enhanced testing environment with dual-view capabilities (Table View and JSON View) for comprehensive guardrail and application testing without code deployment.

  • Batch Processing: Enhanced Python client with support for batch request processing, enabling efficient bulk verification operations.

Enhanced User Experience

  • Real-Time Validation System: Comprehensive status indicators (green/yellow/red) that provide instant feedback on DAG and guardrail readiness, with detailed tooltips explaining any configuration issues.

  • Smart Button States: Context-aware interface that intelligently enables/disables functionality based on current configuration, with helpful hover messages explaining requirements.

  • Interactive Tooltips: Detailed guidance when hovering over disabled elements, showing exactly what needs to be configured (e.g., "Please select: Main LLM, Utility LLM, Embedding Model").

Client Library Improvements

  • Pre-Flight Validation: Jaxon client now validates target applications and guardrails before execution, preventing wasted resources on invalid configurations.

  • Graceful Error Handling: Descriptive error messages for red status targets, with specific details about what needs to be fixed.

  • Enhanced Batch Processing: Improved batch operations that continue processing valid items while clearly identifying and logging validation failures.

Platform Reliability

  • Comprehensive DAG Validation: Deep validation of DAG structures including node configuration, guardrail references, connectivity, and dependency integrity.

  • LLM Requirement Enforcement: Systematic validation of required LLM models for different operations, preventing runtime failures due to missing configurations.

  • Expanded LLM Support: Additional LLM provider integrations and model support for increased flexibility in guardrail configurations.

Platform & Infrastructure

  • Kubernetes Support: Full Kubernetes deployment support with Helm charts for production-scale orchestration and container management.

  • Enhanced UI/UX: Comprehensive admin interface improvements including application management, improved navigation, and streamlined workflow configuration.

  • Advanced Documentation: Restructured documentation with dedicated sections for Applications, Playground, and comprehensive DAG workflow guides.

APIs and Integrations

  • Application API: New application-scoped API endpoints for managing complex verification workflows and DAG configurations.

  • Enhanced Python Client: Improved client library with batch processing capabilities and application workflow execution support.

Release Notes – Version 1.0.0

Core Features

  • Guardrails: Verify LLM outputs with Agentic, Policy Rules (Neurosymbolic), and Manual (Human Review) guardrails.

  • Integrated Logging & Dashboard: Real-time dashboards for monitoring inference.

  • LLM Generalization LLM proxy for decoupled usage of major LLM providers.

APIs and Integrations

  • Python Client: Installable python client for direct integration into customer applications.