Galileo Technologies, Inc.

Streamline AI evaluations with tailored assessments for accurate and reliable outcomes.

Pricing not published

Achieve Real-Time Monitoring and Control Over Your AI Systems.

Pricing not published

Ensure safe and reliable AI with real-time risk mitigation and policy enforcement.

Pricing not published

Proactively detects hidden failures in AI systems to ensure reliability and improve performance.

Pricing not published

Luna-2 ensures reliable AI evaluations and guardrailing for enterprise applications.

Pricing not published

In-environment testing, continuous evaluations, drift detection, and validation workflows that assess models from pre-production through production to ensure policy compliance and performance.

Products: Galileo Evaluate, Galileo Signals, Luna-2

Capture, store, query, and investigate agent traces, prompts, tool calls, and decisions at scale with natural-language search and fast full-text querying for root-cause analysis.

Products: Galileo Observe, Galileo Signals, Galileo Evaluate

Real-time threat detection and context-aware enforcement for AI applications and autonomous agents to prevent prompt injection, data leakage, and unsafe actions with low-latency inline protections and runtime guardrails.

Products: Galileo Protect, Galileo Evaluate, Luna-2

Automated detection of failure patterns from production traces, instant generation of targeted evaluations, institutional memory, and mechanisms for rapid remediation.

Products: Galileo Signals, Galileo Observe, Galileo Evaluate

Galileo Evaluate

Galileo Evaluate is an evaluation platform for generative AI at enterprise scale. It enables teams to evaluate, iterate, monitor, and govern AI applications, offering built-in evaluation capabilities and support for custom evaluators that encode domain knowledge across development lifecycles. Key features include governance through evaluations, real-time monitoring, and support for both prebuilt and customizable metrics.

Pricing not published

• Guides Behavior In Production
  Ensure that agent actions and tool accesses are closely monitored and adjusted based on evaluation results to maintain control and accountability.
• Transforms For Production Guardrails
  Allows seamless transitions from testing environments to production-ready configurations, ensuring AI consistency and reliability are maintained.
• Prebuilt Evaluations Available
  Provides immediate access to essential evaluation tools, enabling faster deployment and adaptability to specific industry requirements.
• Monitors Performance Continuously
  Automatically detects anomalies and potential failures in real-time, allowing for immediate troubleshooting and adjustments.
• Simplifies Integration
  Empowers teams to implement governance without extensive coding, facilitating quick setup and reducing development time.
• Tailors Evaluations To Needs
  Facilitates a targeted approach to evaluation, ensuring results are aligned with specific business or operational objectives.

Galileo Observe

Galileo Observe is the production observability and monitoring solution that provides real-time visibility into AI agent behavior, enabling rapid debugging and performance optimization at scale. Key features include 24/7 monitoring, end-to-end visibility, and real-time alerts to enhance system performance.

Pricing not published

• Ensure Proactive Issue Detection
  Real-time alerts and comprehensive monitoring keep your AI systems operational and efficient.
• Visualize Every Interaction
  Gain complete clarity on how inputs affect outputs to optimize workflows and performance.
• Quickly Respond to Issues
  Instant notifications help teams address problems without delays, reducing downtime.
• Identify Failure Points
  Detailed tracing allows for precise identification of problem areas, streamlining debugging efforts.
• Enhance Monitoring Efficiency
  Customizable insights evolve with your system, ensuring ongoing relevance and optimization.
• Reduce Downtime
  Immediate insights into root causes foster quicker solutions and continuous performance enhancements.

Galileo Protect

Galileo Protect provides guardrails and safeguards to production AI systems, enabling real-time risk mitigation and policy enforcement to prevent harmful or unsafe outputs. Key features include real-time guardrails, customizable policies, and proactive anomaly detection for enhanced security.

Pricing not published

• Blocks Unsafe Outputs
  Automatically prevents harmful responses, ensuring only compliant outputs are generated.
• Customizable Safety Protocols
  Tailor specific rules for your AI applications, addressing unique business requirements.
• Human Oversight
  Ensure a safety net by routing uncertain outputs to human reviewers for evaluation.
• Detect Unknown Issues
  Pinpoint potential problems before they escalate into significant risks.

Galileo Signals

Galileo Signals automatically detects failure patterns in AI agents from production traces, surfacing unknown or emerging issues and enabling rapid evals and remediation. Key features include instant evaluation generation and proactive failure detection to support continuous improvement.

Pricing not published

• Generate Evals Instantly
  Transform identified signals into evaluations with a single click, greatly enhancing workflow efficiency.
• Identify Failures Early
  Detect emerging issues and prevent potential disasters by analyzing 100% of production traces.
• Learn and Improve
  Build institutional knowledge over time to enhance the detection of issues that were previously unknown.
• Track Issue History
  Utilize historical data to accurately assess and manage system reliability.

Luna-2

Luna-2 is Galileo’s production-oriented family of small language models designed for agent reliability and real-time evaluation, available in 3B and 8B parameter sizes and optimized for low latency. Key features include ultra-low latency and an optimized inference engine for consistent performance.

Pricing not published

• Achieve Fast Evaluations
  Luna-2 can process evaluations with ultra-low latency, ensuring optimal performance in time-sensitive AI applications.
• Mitigate Risks Instantly
  With its real-time guardrails, Luna-2 can intercept harmful actions swiftly, enhancing reliability across AI systems.
• Ensure Consistent Performance
  Luna-2 operates on a tailored infrastructure designed for robust AI evaluations, ensuring reliability and speed.
• Choose Optimal Model Size
  Luna-2's flexible configuration allows organizations to select a model that best fits their specific evaluation needs.
• Leverage Extensive Evaluation Metrics
  This architecture allows for efficient monitoring and evaluation across various AI tasks without duplicating resources.