AI Features & Technical Info

Overview

Monte Carlo's product includes AI-powered capabilities designed to enhance functionality and improve efficiency. These features are built with a focus on accuracy, security, and responsible data handling, and are regularly evaluated to ensure they meet quality, compliance, and ethical standards.

LLM (Large Language Model) AI Features

Monte Carlo LLM AI features operate either with LLMs hosted within the Monte Carlo environment or with LLMs hosted within the customer environment.

LLMs Hosted by Monte Carlo

Monitoring Agent: Generates monitor recommendations, suggesting segmentation strategies, translating SQL to business language, and powering generative AI troubleshooting features. This feature includes SQL Query Assistance.

Troubleshooting Agent (Public Preview): Automatically analyzes alerts by testing hundreds of hypotheses, considering data changes, system failures, and code updates, then leverages lineage to pinpoint the most likely root cause.

LLMs Hosted within the Customer Environment

Unstructured Data: Apply sentiment analysis, text classification, or custom prompt-based structuring to freeform text data—without needing to pre-transform data or write SQL.

Agent Observability (Coming Soon): Monitoring, analysis, and understanding of AI systems'.

When an LLM is hosted within the customer environment, it is the responsibility of the customer to ensure the LLM(s) meets appropriate security and privacy requirements.

Approach to AI

We start with clearly defined problems, use high-quality domain data, and design scalable architectures for smooth integration. Models are analyzed, selected, and tested for reliability, with security and compliance built in from the start. Continuous feedback and iteration help the AI improve over time.

AI Quality and Governance

Models are assessed for accuracy and reliability before release, using real-world data and human review. We monitor performance, track for data drift, and conduct regular audits to maintain trust and compliance.

Data Privacy and Opt Out

In all Generative AI use cases within the Monte Carlo platform, no customer data will be used for current or future model training. Monte Carlo customers may disable the Generative AI features for their account at any time by reaching out to Monte Carlo's Customer Support team.