72 docs tagged with "ai"

Send GenAI spans to Coralogix AI Center from .NET using Microsoft.Extensions.AI, which implements the new OpenTelemetry semantic conventions natively.

Deploy a prebuilt observability pack — five Events2Metrics rules and five metric threshold alerts — for monitoring AI applications.

Common questions about AI Center scope, supported integrations, evaluations, pricing, and Code Agents billing impact.

Inspect individual LLM interactions at the span level, view evaluation results and guardrail actions, and trace the full request flow end-to-end in AI Center.

AI Center is a complete platform for AI-powered applications — combining observability, guardrails, evaluations, and AI SPM in one place.

AI Security Posture Management (AI SPM) gives CISOs and security teams a holistic view of AI usage within their organization, enabling them to identify risks and enforce security best practices.

Auto-instrument the Anthropic Claude SDK with OpenTelemetry and send GenAI spans to Coralogix AI Center.

Ask Olly opens the Olly side drawer scoped to the screen you're on — its query, time range, and filters — so you can investigate what you're looking at without describing it first. Available across Coralogix from the page header or with Cmd+I / Ctrl+I.

Let Olly investigate every matching Case automatically, posting a root-cause finding on the Case and routing it to your Case notification destinations.

Olly's autonomous agent runs end-to-end investigations and root cause analysis across your observability data, so you can resolve incidents in minutes instead of hours.

Auto-instrument AWS Bedrock (botocore) with OpenTelemetry and send GenAI spans to Coralogix AI Center using the new semantic conventions.

Classic ML Observability for Coralogix is hosted on the Aporia documentation site.

What's on the Claude dashboard in Code Agents Intelligence — cost, code impact, and per-user activity.

Monitor AI coding agent activity—token usage, costs, tool calls, code changes, repository attribution, and per-user session data—directly in Coralogix.

Copy-pasteable scripts that send OpenTelemetry GenAI spans to Coralogix using the new semantic conventions, in Python, Java, .NET, and Go.

What's on the Codex CLI dashboard in Code Agents Intelligence — tokens, sessions, latency, and per-user activity.

Verified OpenTelemetry GenAI instrumentation libraries for each LLM provider and framework. Coralogix AI Center accepts spans from any instrumentation that emits the conventions.

What's on the Copilot dashboard in Code Agents Intelligence — cost, optimization insights, feature and IDE usage, code impact, and per-user activity.

What's on the Cursor dashboard in Code Agents Intelligence — sessions, lines of code, model and tool usage, and per-user activity.

Create and manage custom evaluation policies in Coralogix AI Center to measure what matters for your specific application.

Define domain-specific guardrail policies using natural language instructions to protect your LLM applications against business-specific risks.

Understand how Olly processes, protects, and governs your data, including architecture, tenant isolation, model provider policies, and compliance frameworks.

Describe what you're looking for in everyday language, and Olly converts your description into a structured DataPrime query.

Olly can return structured results as dynamic tables that can be exported as CSV or JSON.

Enable Olly for a Coralogix team from the Coralogix UI.

Evaluate the quality and safety of your AI applications using prebuilt and custom policies in Coralogix AI Center.

Get clear, natural language explanations for individual log entries, including their significance, impact, and potential causes.

Attach files to a chat message to give Olly additional context for your prompt.

Integrate Coralogix Guardrails with your LLM application to protect against prompt injection, PII leakage, and other security threats.

Instrument your first LLM application with OpenTelemetry and send GenAI spans to Coralogix AI Center in a few minutes.

Integrate Olly with your GitHub repositories to understand code changes, investigate issues, and speed up delivery.

Build GenAI spans by hand in Go with the OpenTelemetry SDK and net/http, and send them to Coralogix AI Center using the new semantic conventions.

Auto-instrument the Google GenAI SDK with OpenTelemetry and forward GenAI spans to Coralogix AI Center through a local collector.

Use the guard() method for full control over multi-turn conversation guardrail evaluation in Coralogix.

Intercept and block harmful, non-compliant, or low-quality AI outputs in real time using Coralogix Guardrails.

Apply ready-to-use Coralogix Guardrails policies for prompt injection, PII detection, and toxicity to protect your LLM applications in real time.

Connect LLM providers and frameworks to Coralogix AI Center for end-to-end visibility into your AI applications.

Build GenAI spans by hand in Java with the OpenTelemetry SDK and send them to Coralogix AI Center using the new semantic conventions.

Ask questions about Coralogix platform features and get real-time answers based on Coralogix documentation.

Understand MCP server limitations.

History of tools added to and removed from the Coralogix MCP server tool catalog.

Let Olly remember helpful facts and preferences from your chats so it personalizes and adds context to future responses, and manage what it stores.

Pick which AI model Olly uses so you control the tradeoff between speed, depth, and cost.

Monitor health, performance, cost, quality, and security posture across all AI applications in AI Center.

Track health, performance, cost, errors, and latency across all your AI applications in AI Center.

The MCP server connects directly to your Coralogix data using OAuth 2.1 with OpenID Connect (OIDC).

Olly is Coralogix's AI-native observability agent that makes observability data fast, accessible, and actionable for everyone.

Use the Olly chat to ask questions, manage conversations, and explore artifacts generated from your observability data.

Frequently asked questions about Olly, the AI-native observability agent.

Olly mini is Coralogix's embedded AI helper — focused, in-context actions like analyzing flame graphs, generating queries, summarizing cases, and explaining logs or errors.

Auto-instrument the OpenAI Agents SDK with OpenTelemetry and send GenAI spans to Coralogix AI Center using the new semantic conventions.

Use the OpenLLMetry (Traceloop) SDK to auto-instrument OpenAI calls and send GenAI spans to Coralogix AI Center.

Send GenAI spans to Coralogix AI Center using the standard OpenTelemetry GenAI semantic conventions — vendor-neutral, no Coralogix-specific SDK required.

Track AI spend across your organization and per application, compare model and user cost, override per-token model pricing, and act on cost-pattern insights detected from live span data.

Bridge your AI tooling and Coralogix telemetry to query observability data, manage alerts and parsing rules, and generate infrastructure as code with the Coralogix MCP server.

Required roles and permissions for accessing and managing AI Center features in Coralogix.

Understand MCP server permissions.

Detect and block personally identifiable information in LLM prompts and responses using the Coralogix Guardrails PII detection policy.

Apply prebuilt evaluation policies in Coralogix AI Center to detect security issues, hallucinations, toxicity, topic violations, and compliance risks, and set a per-eval threshold for what counts as an issue.

Detect and block prompt injection attacks in your LLM applications using the Coralogix Guardrails prompt injection policy.

Build GenAI spans by hand in Python with the OpenTelemetry SDK and raw HTTP — a universal template for any provider or language.

See how code-agent activity splits across the repositories it touches, and classify those repositories as Managed or Unmanaged by configuring Organizations in AI Center.

Define team-wide and personal rules so Olly applies your team's conventions and your individual preferences to every chat.

Schedule Olly to run a prompt automatically — once or on a recurring interval — and review each run from a single page.

Images referenced by ../index.md:

Easily set up Coralogix's MCP server. It is fully remote and does not require any local installation.

Author personal and team-wide skills — by hand or from a chat with Olly — so Olly applies your conventions, workflows, and response style to every chat.

Bring Olly into team conversations on Slack for AI-powered observability collaboration.

Complete inventory of gen_ai.* span attributes consumed by Coralogix AI Center.

Auto-generated catalog of every tool exposed by the Coralogix MCP server.

Detect and block toxic, harmful, or offensive content in LLM prompts and responses using the Coralogix Guardrails toxicity policy.

Monitor team-wide AI Units consumption from Olly and cap it with limits set per user, per user group, or per API key so no one depletes the team quota.