How Do We Know AI Isn’t Lying? The Art of Evaluating LLMs in RAG Systems
🔍 1. Why Evaluating LLM Responses is Hard In classical programming, correctness is binary. Input Expected Result 2 + 2 4 ✔ Correct 2 + 2 5 ✘ Wrong Software is deterministic — same input → same output. LLMs are probabilistic. They generate one of many valid word combinations, like forming sentences from multiple possible synonyms and sentence structures. Example: Prompt: "Explain gravity like I'm 10" Possible responses: Response A Response B Gravity is a force that pulls everything to Earth. Gravity bends space-time causing objects to attract. Both are correct. Which is better? Depends on audience. So evaluation needs to look beyond text similarity. We must check: ✔ Is the answer meaningful? ✔ Is it correct? ✔ Is it easy to understand? ✔ Does it follow prompt intent? Testing LLMs is like grading essays — not checking numeric outputs. 🧠 2. Why RAG Evaluation is Even Harder RAG introduces an additional layer — retrieval. The model no longer answers from memory; it must first read context, then summarise it. Evaluation now has multi-dimensions: Evaluation Layer What we must verify Retrieval Did we fetch the right documents? Understanding Did the model interpret context correctly? Grounding Is the answer based on retrieved data? Generation Quality Is final response complete & clear? A simple story makes this intuitive: Teacher asks student to explain Photosynthesis. Student goes to library → selects a book → reads → writes explanation. We must evaluate: Did they pick the right book? → Retrieval Did they understand the topic? → Reasoning Did they copy facts correctly without inventing? → Faithfulness Is written explanation clear enough for another child to learn from? → Answer Quality One failure → total failure. 🧩 3. Two Types of Evaluation 🔹 Intrinsic Evaluation — Quality of the Response Itself Here we judge the answer, ignoring real-world impact. We check: ✔ Grammar & coherence ✔ Completeness of explanation ✔ No hallucination ✔ Logic flow & clarity ✔ Semantic correctness This is similar to checking how well the essay is written. Even if the result did not solve the real problem, the answer could still look good — that’s why intrinsic alone is not enough. 🔹 Extrinsic Evaluation — Did It Achieve the Goal? This measures task success. If a customer support bot writes a beautifully worded paragraph, but the user still doesn’t get their refund — it failed extrinsically. Examples: System Type Extrinsic Goal Banking RAG Bot Did user get correct KYC procedure? Medical RAG Was advice safe & factual? Legal search assistant Did it return the right section of the law? Technical summariser Did summary capture key meaning? Intrinsic = writing quality. Extrinsic = impact quality. A production-grade RAG system must satisfy both. 📏 4. Core RAG Evaluation Metrics (Explained with Very Simple Analogies) Metric Meaning Analogy Relevance Does answer match question? Ask who invented C++? → model talks about Java ❌ Faithfulness No invented facts Book says started 2004, response says 1990 ❌ Groundedness Answer traceable to sources Claims facts that don’t exist in context ❌ Completeness Covers all parts of question User asks Windows vs Linux → only explains Windows Context Recall / Precision Correct docs retrieved & used Student opens wrong chapter Hallucination Rate Degree of made-up info “Taj Mahal is in London” 😱 Semantic Similarity Meaning-level match “Engine died” = “Car stopped running” 💡 Good evaluation doesn’t check exact wording. It checks meaning + truth + usefulness. 🛠 5. Tools for RAG Evaluation 🔹 1. RAGAS — Foundation for RAG Scoring RAGAS evaluates responses based on: ✔ Faithfulness ✔ Relevance ✔ Context recall ✔ Answer similarity Think of RAGAS as a teacher grading with a rubric. It reads both answer + source documents, then scores based on truthfulness & alignment. 🔹 2. LangChain Evaluators LangChain offers multiple evaluation types: Type What it checks String or regex Basic keyword presence Embedding based Meaning similarity, not text match LLM-as-a-Judge AI evaluates AI (deep reasoning) LangChain = testing toolbox RAGAS = grading framework Together they form a complete QA ecosystem. 🔹 3. PyTest + CI for Automated LLM Testing Instead of manually validating outputs, we automate: Feed preset questions to RAG Capture answers Run RAGAS/LangChain scoring Fail test if hallucination > threshold This brings AI closer to software-engineering discipline. RAG systems stop being experiments — they become testable, trackable, production-grade products. 🚀 6. The Future: LLM-as-a-Judge The future of evaluation is simple: LLMs will evaluate other LLMs. One model writes an answer. Another model checks: ✔ Was it truthful? ✔ Was it relevant? ✔ Did it follow context? This enables: Benefit Why it matters Scalable evaluation No humans needed for every query Continuous improvement Model learns from mistakes Real-time scoring Detect errors before user sees them This is like autopilot for AI systems — not only navigating, but self-correcting mid-flight. And that is where enterprise AI is headed. 🎯 Final Summary Evaluating LLM responses is not checking if strings match. It is checking if the machine: ✔ Understood the question ✔ Retrieved relevant knowledge ✔ Avoided hallucination ✔ Provided complete, meaningful reasoning ✔ Grounded answer in real source text RAG evaluation demands multi-layer validation — retrieval, reasoning, grounding, semantics, safety. Frameworks like RAGAS + LangChain evaluators + PyTest pipelines are shaping the discipline of measurable, reliable AI — pushing LLM-powered RAG from cool demo → trustworthy enterprise intelligence. Useful Resources What is Retrieval-Augmented Generation (RAG) : https://azure.microsoft.com/en-in/resources/cloud-computing-dictionary/what-is-retrieval-augmented-generation-rag/ Retrieval-Augmented Generation concepts (Azure AI) : https://learn.microsoft.com/en-us/azure/ai-services/content-understanding/concepts/retrieval-augmented-generation RAG with Azure AI Search – Overview : https://learn.microsoft.com/en-us/azure/search/retrieval-augmented-generation-overview Evaluate Generative AI Applications (Microsoft Learn – Learning Path) : https://learn.microsoft.com/en-us/training/paths/evaluate-generative-ai-apps/ Evaluate Generative AI Models in Microsoft Foundry Portal : https://learn.microsoft.com/en-us/training/modules/evaluate-models-azure-ai-studio/ RAG Evaluation Metrics (Relevance, Groundedness, Faithfulness) : https://learn.microsoft.com/en-us/azure/ai-foundry/concepts/evaluation-evaluators/rag-evaluators RAGAS – Evaluation Framework for RAG Systems : https://docs.ragas.io/Building Production-Ready, Secure, Observable, AI Agents with Real-Time Voice with Microsoft Foundry
We're excited to announce the general availability of Foundry Agent Service, Observability in Foundry Control Plane, and the Microsoft Foundry portal — plus Voice Live integration with Agent Service in public preview — giving teams a production-ready platform to build, deploy, and operate intelligent AI agents with enterprise-grade security and observability.Generally Available: Evaluations, Monitoring, and Tracing in Microsoft Foundry
If you've shipped an AI agent to production, you've likely run into the same uncomfortable realization: the hard part isn't getting the agent to work - it's keeping it working. Models get updated, prompts get tweaked, retrieval pipelines drift, and user traffic surfaces edge cases that never appeared in your eval suite. Quality isn't something you establish once. It's something you have to continuously measure. Today, we're making that continuous measurement a first-class operational capability. Evaluations, Monitoring, and Tracing in Microsoft Foundry are now generally available through Foundry Control Plane. These aren't standalone tools bolted onto the side of the platform - they're deeply integrated with Azure Monitor, which means AI agent observability now lives in the same operational plane as the rest of your infrastructure. The Problem With Point-in-Time Evaluation Most evaluation workflows are designed around a pre-deployment gate. You build a test dataset, run your evals, review the scores, and ship. That approach has real value - but it has a hard ceiling. In production, agent behavior is a function of many things that change independently of your code: Foundation model updates ship continuously and can shift output style, reasoning patterns, and edge case handling in ways that don't always surface on your benchmark set. Prompt changes can have nonlinear effects downstream, especially in multi-step agentic flows. Retrieval pipeline drift changes what context your agent actually sees at inference time. A document index fresh last month may have stale or subtly different content today. Real-world traffic distribution is never exactly what you sampled for your test set. Production surfaces long-tail inputs that feel obvious in hindsight but were invisible during development. The implication is straightforward: evaluation has to be continuous, not episodic. You need quality signals at development time, at every CI/CD commit, and continuously against live production traffic - all using the same evaluator definitions so results are comparable across environments. That's the core design principle behind Foundry Observability. Continuous Evaluation Across the Full AI Lifecycle Built-In Evaluators Foundry's built-in evaluators cover the most critical quality and safety dimensions for production agent systems: Coherence and Relevance measure whether responses are internally consistent and on-topic relative to the input. These are table-stakes signals for any conversational or task-completion agent. Groundedness is particularly important for RAG-based architectures. It measures whether the model's output is actually supported by the retrieved context - as opposed to plausible-sounding content the model generated from its parametric memory. Groundedness failures are a leading indicator of hallucination risk in production, and they're often invisible to human reviewers at scale. Retrieval Quality evaluates the retrieval step independently from generation. Groundedness failures can originate in two places: the model may be ignoring good context, or the retrieval pipeline may not be surfacing relevant context in the first place. Splitting these signals makes it much easier to pinpoint root cause. Safety and Policy Alignment evaluates whether outputs meet your deployment's policy requirements - content safety, topic restrictions, response format compliance, and similar constraints. These evaluators are designed to run at every stage of the AI lifecycle: Local development - run evals inline as you iterate on prompts, retrieval config, or orchestration logic CI/CD pipelines - gate every commit against your quality baselines; catch regressions before they reach production Production traffic monitoring - continuously evaluate sampled live traffic and surface trends over time Because the evaluators are identical across all three contexts, a score in CI means the same thing as a score in production monitoring. See the Practical Guide to Evaluations and the Built-in Evaluators Reference for a deeper walkthrough. Custom Evaluators - Encoding Your Own Definition of Quality Built-in evaluators cover common signals well, but production agents often need to satisfy criteria specific to a domain, regulatory environment, or internal standard. Foundry supports two types of custom evaluators (currently in public preview): LLM-as-a-Judge evaluators let you configure a prompt and grading rubric, then use a language model to apply that rubric to your agent's outputs. This is the right approach for quality dimensions that require reasoning or contextual judgment - whether a response appropriately acknowledges uncertainty, whether a customer-facing message matches your brand tone, or whether a clinical summary meets documentation standards. You write a judge prompt with a scoring scale (e.g., 1–5 with criteria for each level) that evaluates a given {input} / {response} pair. Foundry runs this at scale and aggregates scores into your dashboards alongside built-in results. Code-based evaluators are Python functions that implement any evaluation logic you can express programmatically - regex matching, schema validation, business rule checks, compliance assertions, or calls to external systems. If your organization has documented policies about what a valid agent response looks like, you can encode those policies directly into your evaluation pipeline. Custom and built-in evaluators compose naturally - running against the same traffic, producing results in the same schema, feeding into the same dashboards and alert rules. Monitoring and Alerting - AI Quality as an Operational Signal All observability data produced by Foundry - evaluation results, traces, latency, token usage, and quality metrics - is published directly to Azure Monitor. This is where the integration pays off for teams already on Azure. What this enables that siloed AI monitoring tools can't: Cross-stack correlation. When your groundedness score drops, is it a model update, a retrieval pipeline issue, or an infrastructure problem affecting latency? With AI quality signals and infrastructure telemetry in the same Azure Monitor Application Insights workspace, you can answer that in minutes rather than hours of manual correlation across disconnected systems. Unified alerting. Configure Azure Monitor alert rules on any evaluation metric - trigger a PagerDuty incident when groundedness drops below threshold, send a Teams notification when safety violations spike, or create automated runbook responses when retrieval quality degrades. These are the same alert mechanisms your SRE team already uses. Enterprise governance by default. Azure Monitor's RBAC, retention policies, diagnostic settings, and audit logging apply automatically to all AI observability data. You inherit the governance framework your organization has already built and approved. Grafana and existing dashboards. If your team uses Azure Managed Grafana, evaluation metrics can flow into existing dashboards alongside your other operational metrics - a single pane of glass for application health, infrastructure performance, and AI agent quality. The Agent Monitoring Dashboard in the Foundry portal provides an AI-native view out of the box - evaluation metric trends, safety threshold status, quality score distributions, and latency breakdowns. Everything in that dashboard is backed by Azure Monitor data, so SRE teams can always drill deeper. End-to-End Tracing: From Quality Signal to Root Cause A groundedness score tells you something is wrong. A trace tells you exactly where the failure occurred and what the agent actually did. Foundry provides OpenTelemetry-based distributed tracing that follows each request through your entire agent system: model calls, tool invocations, retrieval steps, orchestration logic, and cross-agent handoffs. Traces capture the full execution path - inputs, outputs, latency at each step, tool call parameters and responses, and token usage. The key design decision: evaluation results are linked directly to traces. When you see a low groundedness score in your monitoring dashboard, you navigate directly to the specific trace that produced it - no manual timestamp correlation, no separate trace ID lookup. The connection is made automatically. Foundry auto-collects traces across the frameworks your agents are likely already built on: Microsoft Agent Framework Semantic Kernel LangChain and LangGraph OpenAI Agents SDK For custom or less common orchestration frameworks, the Azure Monitor OpenTelemetry Distro provides an instrumentation path. Microsoft is also contributing upstream to the OpenTelemetry project - working with Cisco Outshift, we've contributed semantic conventions for multi-agent trace correlation, standardizing how agent identity, task context, and cross-agent handoffs are represented in OTel spans. Note: Tracing is currently in public preview, with GA shipping by end of March. Prompt Optimizer (Public Preview) One persistent friction point in agent development is the iteration loop between writing prompts and measuring their effect. You make a change, run your evals, look at the delta, try to infer what about the change mattered, and repeat. Prompt Optimizer tightens this loop. It analyzes your existing prompt and applies structured prompt engineering techniques - clarifying ambiguous instructions, improving formatting for model comprehension, restructuring few-shot examples, making implicit constraints explicit - with paragraph-level explanations for every change it makes. The transparency is deliberate. Rather than producing a black-box "optimized" prompt, it shows you exactly what it changed and why. You can add constraints, trigger another optimization pass, and iterate until satisfied. When you're done, apply it with one click. The value compounds alongside continuous evaluation: run your eval suite against the current prompt, optimize, run evals again, see the measured improvement. That feedback loop - optimize, measure, optimize - is the closest thing to a systematic approach to prompt engineering that currently exists. What Makes our Approach to Observability Different There are other evaluation and observability tools in the AI ecosystem. The differentiation in Foundry's approach comes down to specific architectural choices: Unified lifecycle coverage, not just pre-deployment testing. Most existing evaluation tools are designed for offline, pre-deployment use. Foundry's evaluators run in the same form at development time, in CI/CD, and against live production traffic. Your quality metrics are actually comparable across the lifecycle - you can tell whether production quality matches what you saw in testing, rather than operating two separate measurement systems that can't be compared. No separate observability silo. Publishing all observability data to Azure Monitor means you don't operate a separate system for AI quality alongside your existing infrastructure monitoring. AI incidents route through your existing on-call rotations. AI quality data is subject to the same retention and compliance controls as the rest of your telemetry. Framework-agnostic tracing. Auto-instrumentation across Semantic Kernel, LangChain, LangGraph, and the OpenAI Agents SDK means you're not locked into a specific orchestration framework. The OpenTelemetry foundation means trace data is portable to any compatible backend, protecting your investment as the tooling landscape evolves. Composable evaluators. Built-in and custom evaluators run in the same pipeline, against the same traffic, producing results in the same schema, feeding into the same dashboards and alert rules. You don't choose between generic coverage and domain-specific precision - you get both. Evaluation linked to traces. Most systems treat evaluation and tracing as separate concerns. Foundry treats them as two views of the same event - closing the loop between detecting a quality problem and diagnosing it. Getting Started If you're building agents on Microsoft Foundry, or using Semantic Kernel, LangChain, LangGraph, or the OpenAI Agents SDK and want to add production observability, the entry point is Foundry Control Plane. Try it You'll need a Foundry project with an agent and an Azure OpenAI deployment. Enable observability by navigating to Foundry Control Plane and connecting your Azure Monitor workspace. Then walk through the Practical Guide to Evaluations, explore the Built-in Evaluators Reference, and set up end-to-end tracing for your agents.Foundry Agent Service at Ignite 2025: Simple to Build. Powerful to Deploy. Trusted to Operate.
The upgraded Foundry Agent Service delivers a unified, simplified platform with managed hosting, built-in memory, tool catalogs, and seamless integration with Microsoft Agent Framework. Developers can now deploy agents faster and more securely, leveraging one-click publishing to Microsoft 365 and advanced governance features for streamlined enterprise AI operations.Observability for Multi-Agent Systems with Microsoft Agent Framework and Azure AI Foundry
Agentic applications are revolutionizing enterprise automation, but their dynamic toolchains and latent reasoning make them notoriously hard to operate. In this post, you'll learn how to instrument a Microsoft Agent Framework–based service with OpenTelemetry, ship traces to Azure AI Foundry observability, and adopt a practical workflow to debug, evaluate, and improve multi-agent behavior in production. We'll show how to wire spans around reasoning steps and tool calls (OpenAPI / MCP), enabling deep visibility into your agentic workflows. Who Should Read This? Developers building agents with Microsoft Agent Framework (MAF) in .NET or Python Architects/SREs seeking enterprise-grade visibility, governance, and reliability for deployments on Azure AI Foundry Why Observability Is Non-Negotiable for Agents Traditional logs fall short for agentic systems: Reasoning and routing (which tool? which doc?) are opaque without explicit spans/events Failures often occur between components (e.g., retrieval mismatch, tool schema drift) Without traces across agents ⇄ tools ⇄ data stores, you can't reproduce or evaluate behavior Microsoft has introduced multi-agent observability patterns and OpenTelemetry (OTel) conventions that unify traces across Agent Framework, Foundry, and popular stacks—so you can see one coherent timeline for each task. Reference Architecture Key Capabilities Agent orchestration & deployment via Microsoft Agent Framework Model access using Foundry’s OpenAI-compatible endpoint OpenTelemetry for traces/spans + attributes (agent, tool, retrieval, latency, tokens) Step-by-Step Implementation Assumption: This article uses Azure Monitor (via Application Insights) as the OpenTelemetry exporter, but you can configure other supported exporters in the same way. Prerequisites .NET 8 SDK or later Azure OpenAI service (endpoint, API key, deployed model) Application Insights and Grafana Create an Agent with OpenTelemetry (ASP.NET Core or Console App) Install required packages: dotnet add package Azure.AI.OpenAI dotnet add package Azure.Monitor.OpenTelemetry.Exporter dotnet add package Microsoft.Agents.AI.OpenAI dotnet add package Microsoft.Extensions.Logging dotnet add package OpenTelemetry dotnet add package OpenTelemetry.Trace dotnet add package OpenTelemetry.Metrics dotnet add package OpenTelemetry.Extensions.Hosting dotnet add package OpenTelemetry.Instrumentation.Http Setup environment variables: AZURE_OPENAI_ENDPOINT: https://<your_service_name>.openai.azure.com/ AZURE_OPENAI_API_KEY: <your_azure_openai_apikey> APPLICATIONINSIGHTS_CONNECTION_STRING: <your_application_insights_connectionstring_for_azuremonitor_exporter> Configure tracing once at startup: var applicationInsightsConnectionString = Environment.GetEnvironmentVariable("APPLICATIONINSIGHTS_CONNECTION_STRING"); // Create a resource describing the service var resource = ResourceBuilder.CreateDefault() .AddService(serviceName: ServiceName) .AddAttributes(new Dictionary<string, object> { ["deployment.environment"] = "development", ["service.instance.id"] = Environment.MachineName }) .Build(); // Setup OpenTelemetry TracerProvider var traceProvider = Sdk.CreateTracerProviderBuilder() .SetResourceBuilder(ResourceBuilder.CreateDefault().AddService(ServiceName)) .AddSource(SourceName) .AddSource("Microsoft.Agents.AI") .AddHttpClientInstrumentation() .AddAzureMonitorTraceExporter(options => { options.ConnectionString = applicationInsightsConnectionString; }) .Build(); // Setup OpenTelemetry MeterProvider var meterProvider = Sdk.CreateMeterProviderBuilder() .SetResourceBuilder(ResourceBuilder.CreateDefault().AddService(ServiceName)) .AddMeter(SourceName) .AddAzureMonitorMetricExporter(options => { options.ConnectionString = applicationInsightsConnectionString; }) .Build(); // Configure DI and OpenTelemetry var serviceCollection = new ServiceCollection(); // Setup Logging with OpenTelemetry and Application Insights serviceCollection.AddLogging(loggingBuilder => { loggingBuilder.SetMinimumLevel(LogLevel.Debug); loggingBuilder.AddOpenTelemetry(options => { options.SetResourceBuilder(ResourceBuilder.CreateDefault().AddService(ServiceName)); options.IncludeScopes = true; options.IncludeFormattedMessage = true; options.AddAzureMonitorLogExporter(exporterOptions => { exporterOptions.ConnectionString = applicationInsightsConnectionString; }); }); loggingBuilder.AddApplicationInsights( configureTelemetryConfiguration: (config) => { config.ConnectionString = Environment.GetEnvironmentVariable("APPLICATIONINSIGHTS_CONNECTION_STRING"); }, configureApplicationInsightsLoggerOptions: options => { options.TrackExceptionsAsExceptionTelemetry = true; options.IncludeScopes = true; }); }); Configure custom metrics and activity source for tracing: using var activitySource = new ActivitySource(SourceName); using var meter = new Meter(SourceName); // Create custom metrics var interactionCounter = meter.CreateCounter<long>("chat_interactions_total", description: "Total number of chat interactions"); var responseTimeHistogram = meter.CreateHistogram<double>("chat_response_time_ms", description: "Chat response time in milliseconds"); 2. Wire-up the AI Agent: // Create OpenAI client var endpoint = Environment.GetEnvironmentVariable("AZURE_OPENAI_ENDPOINT"); var apiKey = Environment.GetEnvironmentVariable("AZURE_OPENAI_API_KEY"); var deploymentName = "gpt-4o-mini"; using var client = new AzureOpenAIClient(new Uri(endpoint), new AzureKeyCredential(apiKey)) .GetChatClient(deploymentName) .AsIChatClient() .AsBuilder() .UseOpenTelemetry(sourceName: SourceName, configure: (cfg) => cfg.EnableSensitiveData = true) .Build(); logger.LogInformation("Creating Agent with OpenTelemetry instrumentation"); // Create AI Agent var agent = new ChatClientAgent( client, name: "AgentObservabilityDemo", instructions: "You are a helpful assistant that provides concise and informative responses.") .AsBuilder() .UseOpenTelemetry(SourceName, configure: (cfg) => cfg.EnableSensitiveData = true) .Build(); var thread = agent.GetNewThread(); logger.LogInformation("Agent created successfully with ID: {AgentId}", agent.Id); 3. Instrument Agent logic with semantic attributes and call OpenAI-compatible API: // Create a parent span for the entire agent session using var sessionActivity = activitySource.StartActivity("Agent Session"); Console.WriteLine($"Trace ID: {sessionActivity?.TraceId} "); var sessionId = Guid.NewGuid().ToString("N"); sessionActivity? .SetTag("agent.name", "AgentObservabilityDemo") .SetTag("session.id", sessionId) .SetTag("session.start_time", DateTimeOffset.UtcNow.ToString("O")); logger.LogInformation("Starting agent session with ID: {SessionId}", sessionId); using (logger.BeginScope(new Dictionary<string, object> { ["SessionId"] = sessionId, ["AgentName"] = "AgentObservabilityDemo" })) { var interactionCount = 0; while (true) { Console.Write("You (or 'exit' to quit): "); var input = Console.ReadLine(); if (string.IsNullOrWhiteSpace(input) || input.Equals("exit", StringComparison.OrdinalIgnoreCase)) { logger.LogInformation("User requested to exit the session"); break; } interactionCount++; logger.LogInformation("Processing interaction #{InteractionCount}", interactionCount); // Create a child span for each individual interaction using var activity = activitySource.StartActivity("Agent Interaction"); activity? .SetTag("user.input", input) .SetTag("agent.name", "AgentObservabilityDemo") .SetTag("interaction.number", interactionCount); var stopwatch = Stopwatch.StartNew(); try { logger.LogInformation("Starting agent execution for interaction #{InteractionCount}", interactionCount); var response = await agent.RunAsync(input); Console.WriteLine($"Agent: {response}"); Console.WriteLine(); stopwatch.Stop(); var responseTimeMs = stopwatch.Elapsed.TotalMilliseconds; // Record metrics interactionCounter.Add(1, new KeyValuePair<string, object?>("status", "success")); responseTimeHistogram.Record(responseTimeMs, new KeyValuePair<string, object?>("status", "success")); activity?.SetTag("interaction.status", "success"); logger.LogInformation("Agent interaction #{InteractionNumber} completed successfully in {ResponseTime:F2} seconds", interactionCount, responseTimeMs); } catch (Exception ex) { Console.WriteLine($"Error: {ex.Message}"); Console.WriteLine(); stopwatch.Stop(); var responseTimeMs = stopwatch.Elapsed.TotalSeconds; // Record error metrics interactionCounter.Add(1, new KeyValuePair<string, object?>("status", "error")); responseTimeHistogram.Record(responseTimeMs, new KeyValuePair<string, object?>("status", "error")); activity? .SetTag("response.success", false) .SetTag("error.message", ex.Message) .SetStatus(ActivityStatusCode.Error, ex.Message); logger.LogError(ex, "Agent interaction #{InteractionNumber} failed after {ResponseTime:F2} seconds: {ErrorMessage}", interactionCount, responseTimeMs, ex.Message); } } // Add session summary to the parent span sessionActivity? .SetTag("session.total_interactions", interactionCount) .SetTag("session.end_time", DateTimeOffset.UtcNow.ToString("O")); logger.LogInformation("Agent session completed. Total interactions: {TotalInteractions}", interactionCount); Azure Monitor dashboard Once you run the agent and generate some traffic, your dashboard in Azure Monitor will be populated as shown below: You can drill down to specific service / activity source / spans by applying relevant filters: Key Features Demonstrated OpenTelemetry instrumentation with Microsoft Agent framework Custom metrics for user interactions End-to-end Telemetry correlation Real time telemetry visualization along with metrics and logging interactions Further reading Introducing Microsoft Agent Framework Azure AI Foundry docs OpenTelemetry Aspire Demo with Azure OpenAIWant Safer, Smarter AI? Start with Observability in Azure AI Foundry
Observability in Azure AI: From Black Box to Transparent Intelligence If you are an AI developer or an engineer, you can benefit from Azure AI observability by gaining deep visibility into agent behavior, enabling them to trace decisions, evaluate response quality, and integrate automated testing into their workflows. This empowers you to build safer, more reliable GenAI applications. Responsible AI and compliance teams use observability tools to ensure transparency and accountability, leveraging audit logs, policy mapping, and risk scoring. These capabilities help organizations align AI development with ethical standards and regulatory requirements. Understanding Observability Imagine you're building a customer support chatbot using Azure AI. It’s designed to answer billing questions, troubleshoot issues, and escalate complex cases to human agents. Everything works well in testing—but once deployed, users start reporting confusing answers and slow response times. Without observability, you’re flying blind. You don’t know: Which queries are failing. Why the chatbot is choosing certain responses. Whether it's escalating too often or not enough. How latency and cost are trending over time. Enter Observability: With Azure AI Foundry and Azure Monitor, you can: Trace every interaction: See the full reasoning path the chatbot takes—from user input to model invocation to tool calls. Evaluate response quality: Automatically assess whether answers are grounded, fluent, and relevant. Monitor performance: Track latency, throughput, and cost per interaction. Detect anomalies: Use Azure Monitor’s ML-powered diagnostics to spot unusual patterns. Improve continuously: Feed evaluation results back into your CI/CD pipeline to refine the chatbot with every release. This is observability in action: turning opaque AI behavior into transparent, actionable insights. It’s not just about fixing bugs—it’s about building AI you can trust. Next, let’s understand more about observability: What Is Observability in Azure AI? Observability in Azure AI refers to the ability to monitor, evaluate, and govern AI agents and applications across their lifecycle—from model selection to production deployment. It’s not just about uptime or logs anymore. It’s about trust, safety, performance, cost, and compliance. Observability aligned with the end-to-end AI application development workflow. Image source: Microsoft Learn Key Components and Capabilities Azure AI Foundry Observability Built-in observability for agentic workflows. Tracks metrics like performance, quality, cost, safety, relevance, and “groundedness” in real time. Enables tracing of agent interactions and data lineage. Supports alerts for risky or off-policy responses and integrates with partner governance platforms. Find details on Observability here: Observability in Generative AI with Azure AI Foundry - Azure AI Foundry | Microsoft Learn AI Red Teaming (PyRIT Integration) Scans agents for safety vulnerability. Evaluates attack success rates across categories like hate, violence, sexual content, and l more. Generates scorecards and logs results in the Foundry portal. Find details here: AI Red Teaming Agent - Azure AI Foundry | Microsoft Learn Image source: Microsoft Learn CI/CD Integration GitHub Actions and Azure DevOps workflows automate evaluations. Continuous monitoring and regression detection during development Azure Monitor + Azure BRAIN Uses ML and LLMs for anomaly detection, forecasting, and root cause analysis. Offers multi-tier log storage (Gold, Silver, Bronze) with unified KQL query experience. Integrates with Azure Copilot for diagnostics and optimization. Open Telemetry Extensions Azure is extending OTel with agent-specific entities like AgentRun, ToolCall, Eval, and ModelInvocation. Enables fleet-scale dashboards and semantic tracing for GenAI workloads. Observability as a First-Class Citizen in Azure AI Foundry In Azure AI Foundry, observability isn’t bolted on—it’s built in. The platform treats observability as a first-class capability, essential for building trustworthy, scalable, and responsible AI systems. Image source: Microsoft Learn What Does This Mean in Practice? Semantic Tracing for Agents Azure AI Foundry enables intelligent agents to perform tasks using AgentRun, ToolCall, and ModelInvocation. AgentRun manages the entire lifecycle of an agent's execution, from input processing to output generation. ToolCall allows agents to invoke external tools or APIs for specific tasks, like fetching data or performing calculations. ModelInvocation lets agents directly use AI models for advanced tasks, such as sentiment analysis or image recognition. Together, these components create adaptable agents capable of handling complex workflows efficiently. Integrated Evaluation Framework Developers can continuously assess agent responses for quality, safety, and relevance using built-in evaluators. These can be run manually or automatically via CI/CD pipelines, enabling fast iteration and regression detection. Governance and Risk Management Observability data feeds directly into governance workflows. Azure AI Foundry supports policy mapping, risk scoring, and audit logging, helping teams meet compliance requirements while maintaining agility. Feedback Loop for Continuous Improvement Observability isn’t just about watching—it’s about learning. Azure AI Foundry enables teams to use telemetry and evaluation data to refine agents, improve performance, and reduce risk over time. Now, Build AI You Can Trust Observability isn’t just a technical feature—it’s the foundation of responsible AI. Whether you're building copilots, deploying GenAI agents, or modernizing enterprise workflows, Azure AI Foundry and Azure Monitor give you the tools to trace, evaluate, and improve every decision your AI makes. Now is the time to move beyond black-box models and embrace transparency, safety, and performance at scale. Start integrating observability into your AI workflows and unlock the full potential of your agents—with confidence. Read more here: Plans | Microsoft Learn Observability and Continuous Improvement - Training | Microsoft Learn Observability in Generative AI with Azure AI Foundry - Azure AI Foundry | Microsoft Learn About the Author Priyanka is a Technical Trainer at Microsoft USA with over 15 years of experience as a Microsoft Certified Trainer. She has a profound passion for learning and sharing knowledge across various domains. Priyanka excels in delivering training sessions, proctoring exams, and upskilling Microsoft Partners and Customers. She has significantly contributed to AI and Data-related courseware, exams, and high-profile events such as Microsoft Ignite, Microsoft Learn Live Shows, MCT Community AI Readiness, and Women in Cloud Skills Ready. Furthermore, she supports initiatives like “Code Without Barrier” and “Women in Azure AI,” contributing to AI Skills enhancements. Her primary areas of expertise include courses on Development, Data, and AI. In addition to maintaining and acquiring new certifications in Data and AI, she has also guided learners and enthusiasts on their educational journeys. Priyanka is an active member of the Microsoft Tech community, where she reviews and writes blogs focusing on Data and AI. #SkilledByMTT #MSLearn #MTTBloggingGroupThe Future of AI: Reduce AI Provisioning Effort - Jumpstart your solutions with AI App Templates
In the previous post, we introduced Contoso Chat – an open-source RAG-based retail chat sample for Azure AI Foundry, that serves as both an AI App template (for builders) and the basis for a hands-on workshop (for learners). And we briefly talked about five stages in the developer workflow (provision, setup, ideate, evaluate, deploy) that take them from the initial prompt to a deployed product. But how can that sample help you build your app? The answer lies in developer tools and AI App templates that jumpstart productivity by giving you a fast start and a solid foundation to build on. In this post, we answer that question with a closer look at Azure AI App templates - what they are, and how we can jumpstart our productivity with a reuse-and-extend approach that builds on open-source samples for core application architectures.The Future of AI: Harnessing AI agents for Customer Engagements
Discover how AI-powered agents are revolutionizing customer engagement—enhancing real-time support, automating workflows, and empowering human professionals with intelligent orchestration. Explore the future of AI-driven service, including Customer Assist created with Azure AI Foundry.
AI reports: Improve AI governance and GenAIOps with consistent documentation
AI reports are designed to help organizations improve cross-functional observability, collaboration, and governance when developing, deploying, and operating generative AI applications and fine-tuned or custom models. These reports support AI governance best practices by helping developers document the purpose of their AI model or application, its features, potential risks or harms, and applied mitigations, so that cross-functional teams can track and assess production-readiness throughout the AI development lifecycle and then monitor it in production. Starting in December, AI reports will be available in private preview in a US and EU Azure region for Azure AI Foundry customers. To request access to the private preview of AI reports, please complete the Interest Form. Furthermore, we are excited to announce new collaborations with Credo AI and Saidot to support customers’ end-to-end AI governance. By integrating the best of Azure AI with innovative and industry-leading AI governance solutions, we hope to provide our customers with choice and help empower greater cross-functional collaboration to align AI solutions with their own principles and regulatory requirements. Building on learnings at Microsoft Microsoft’s approach for governing generative AI applications builds on our Responsible AI Standard and the National Institute of Standards and Technology’s AI Risk Management Framework. This approach requires teams to map, measure, and manage risks for generative applications throughout their development cycle. A core asset of the first—and iterative—map phase is the Responsible AI Impact Assessment. These assessments help identify potential risks and their associated harms, as well as mitigations to address them. As development of an AI system progresses, additional iterations can help development teams document their progress in risk mitigation and allow experts to review the evaluations and mitigations and make further recommendations or requirements before products are launched. Post-deployment, these assessments become a source of truth for ongoing governance and audits, and help guide how to monitor the application in production. You can learn more about Microsoft’s approach to AI governance in our Responsible AI Transparency Report and find a Responsible AI Impact Assessment Guide and example template on our website. How AI reports support AI impact assessments and GenAIOps AI reports can help organizations govern their GenAI models and applications by making it easier for developers to provide the information needed for cross-functional teams to assess production-readiness throughout the GenAIOps lifecycle. Developers will be able to assemble key project details, such as the intended business use case, potential risks and harms, model card, model endpoint configuration, content safety filter settings, and evaluation results into a unified AI report from within their development environment. Teams can then publish these reports to a central dashboard in the Azure AI Foundry portal, where business leaders can track, review, update, and assess reports from across their organization. Users can also export AI reports in PDF and industry-standard SPDX 3.0 AI BOM formats, for integration into existing GRC workflows. These reports can then be used by the development team, their business leaders, and AI, data, and other risk professionals to determine if an AI model or application is fit for purpose and ready for production as part of their AI impact assessment processes. Being versioned assets, AI reports can also help organizations build a consistent bridge across experimentation, evaluation, and GenAIOps by documenting what metrics were evaluated, what will be monitored in production, and the thresholds that will be used to flag an issue for incident response. For even greater control, organizations can choose to implement a release gate or policy as part of their GenAIOps that validates whether an AI report has been reviewed and approved for production. Key benefits of these capabilities include: Observability: Provide cross-functional teams with a shared view of AI models and applications in development, in review, and in production, including how these projects perform in key quality and safety evaluations. Collaboration: Enable consistent information-sharing between GRC, development, and operational teams using a consistent and extensible AI report template, accelerating feedback loops and minimizing non-coding time for developers. Governance: Facilitate responsible AI development across the GenAIOps lifecycle, reinforcing consistent standards, practices, and accountability as projects evolve or expand over time. Build production-ready GenAI apps with Azure AI Foundry If you are interested in testing AI reports and providing feedback to the product team, please request access to the private preview by completing the Interest Form. Want to learn more about building trustworthy GenAI applications with Azure AI? Here’s more guidance and exciting announcements to support your GenAIOps and governance workflows from Microsoft Ignite: Learn about new GenAI evaluation capabilities in Azure AI Foundry Learn about new GenAI monitoring capabilities in Azure AI Foundry Learn about new IT governance capabilities in Azure AI Foundry Whether you’re joining in person or online, we can’t wait to see you at Microsoft Ignite 2024. We’ll share the latest from Azure AI and go deeper into capabilities that support trustworthy AI with these sessions: Keynote: Microsoft Ignite Keynote Breakout: Trustworthy AI: Future trends and best practices Breakout: Trustworthy AI: Advanced AI risk evaluation and mitigation Demo: Simulate, evaluate, and improve GenAI outputs with Azure AI Foundry Demo: Track and manage GenAI app risks with AI reports in Azure AI Foundry We’ll also be available for questions in the Connection Hub on Level 3, where you can find “ask the expert” stations for Azure AI and Trustworthy AI.
Continuously monitor your GenAI application with Azure AI Foundry and Azure Monitor
Now, Azure AI Foundry and Azure Monitor seamlessly integrate to enable ongoing, comprehensive monitoring of your GenAI application's performance from various perspectives, including token usage, operational metrics (e.g. latency and request count), and the quality and safety of generated outputs. With online evaluation, now available in public preview, you can continuously assess your application's outputs, regardless of its deployment or orchestration framework, using built-in or custom evaluation metrics. This approach can help organizations identify and address security, quality, and safety issues in both pre-production and post-production phases of the enterprise GenAIOps lifecycle. Additionally, online evaluations integrate seamlessly with new tracing capabilities in Azure AI Foundry, now available in public preview, as well as Azure Monitor Application Insights. Tying it all together, Azure Monitor enables you to create custom monitoring dashboards, visualize evaluation results over time, and set up alerts for advanced monitoring and incident response. Let’s dive into how all these monitoring capabilities fit together to help you be successful when building enterprise-ready GenAI applications. Observability and the enterprise GenAIOps lifecycle The generative AI operations (GenAIOps) lifecycle is a dynamic development process that spans all the way from ideation to operationalization. It involves choosing the right base model(s) for your application, testing and making changes to the flow, and deploying your application to production. Throughout this process, you can evaluate your application’s performance iteratively and continuously. This practice can help you identify and mitigate issues early and optimize performance as you go, helping ensure your application performs as expected. You can use the built-in evaluation capabilities in Azure AI Foundry, which now include remote evaluation and continuous online evaluation, to support end-to-end observability into your app’s performance throughout the GenAIOps lifecycle. Online evaluation can be used in many different application development scenarios, including: Automated testing of application variants. Integration into DevOps CI/CD pipelines. Regularly assessing an application’s responses for key quality metrics (e.g. groundedness, coherence, recall). Quickly responding to risky or inappropriate outputs that may arise during real-world use (e.g. containing violent, hateful, or sexual content) Production application monitoring and observability with Azure Monitor Application Insights. Now, let explore how you can use tracing for your application to begin your observability journey. Gain deeper insight into your GenAI application's processes with tracing Tracing enables comprehensive monitoring and deeper analysis of your GenAI application's execution. This functionality allows you to trace the process from input to output, review intermediate results, and measure execution times. Additionally, detailed logs for each function call in your workflow are accessible. You can inspect parameters, metrics, and outputs of each AI model utilized, which facilitates debugging and optimization of your application while providing deeper insights into the functioning and outputs of the AI models. The Azure AI Foundry SDK supports tracing to various endpoints, including local viewers, Azure AI Foundry, and Azure Monitor Application Insights. Learn more about new tracing capabilities in Azure AI Foundry. Continuously measure the quality and safety of generated outputs with online evaluation With online evaluation, now available in public preview, you can continuously evaluate your collected trace data for troubleshooting, monitoring, and debugging purposes. Online evaluation with Azure AI Foundry offers the following capabilities: Integration between Azure AI services and Azure Monitor Application Insights Monitor any deployed application, agnostic of deployment method or orchestration framework Support for trace data logged via the Azure AI Foundry SDK or a logging API of your choice Support for built-in and custom evaluation metrics via the Azure AI Foundry SDK Can be used to monitor your application during all stages of the GenAIOps lifecycle To get started with online evaluation, please review the documentation and code samples. Monitor your app in production with Azure AI Foundry and Azure Monitor Azure Monitor Application Insights excels in application performance monitoring (APM) for live web applications, providing many experiences to help enhance the performance, reliability, and quality of your applications. Once you’ve started collecting data for your GenAI application, you can access an out-of-the-box dashboard view to help you get started with monitoring key metrics for your application directly from your Azure AI project. Insights are surfaced to you via an Azure Monitor workbook that is linked to your Azure AI project, helping you quickly observe trends for key metrics, such as token consumption, user feedback, and evaluations. You can customize this workbook and add tiles for additional metrics or insights based on your business needs. You can also share it with your team so they can get the latest insights as well. Build enterprise-ready GenAI apps with Azure AI Foundry Ready to learn more? Here are other exciting announcements from Microsoft Ignite to support your GenAIOps workflows: New tracing and debugging capabilities to drive continuous improvement New ways to evaluate models and applications in pre-production New ways to document and share evaluation results with business stakeholders Whether you’re joining in person or online, we can’t wait to see you at Microsoft Ignite 2024. We’ll share the latest from Azure AI and go deeper into best practices for GenAIOps with these breakout sessions: Multi-agentic GenAIOps from prototype to production with dev tools Trustworthy AI: Advanced risk evaluation and mitigation Azure AI and the dev toolchain you need to infuse AI in all your apps
