Evaluate before you ship: introducing the Voice Live Evaluation Harness
You've built a voice agent on Azure Voice Live. It demos beautifully. Then a teammate asks the question that keeps every voice-agent team up at night: "How do we know it's actually good — across 200 customer calls, not the three we just listened to?" Until today, the honest answer was: put on headphones. Manual listening. Subjective scoring in a spreadsheet. No baseline, no regression signal, no way to defend a model swap with data. We're releasing the Voice Live Evaluation Harness to change that. It's an open-source, deployable evaluation pipeline that runs pre-recorded multi-turn audio through your Voice Live agent and scores every turn with the same evaluators built into Microsoft Foundry — automatically, repeatably, and in parallel. TL;DR Two flavors, one repo. Run the CLI harness locally against a Foundry project for fast iteration, or deploy the evaluation agent into your Azure subscription with the Azure Developer CLI (azd) for a fully-hosted evaluation backend. 13 built-in evaluators score every turn — intent resolution, task adherence, task completion, response completeness, tool-call accuracy, groundedness, and more — viewable per-turn and in aggregate inside the Foundry portal. Supports the three Voice Live modes you actually ship in — Semantic VAD, Push-to-Talk, and Foundry Agent mode — including multi-turn conversations with tool calls and grounding. Grows with your agent. Start with the sample datasets, then layer in audio collected from user testing and production traffic so your evaluation set matures alongside the agent. 🔗 Repo: microsoft-foundry/voicelive-evaluation · Docs: Evaluate Voice Live agents (preview) Why systematic evaluation matters for voice agents Text agents have a mature evaluation story. Voice agents don't — and the gaps actually matter more, because every voice failure happens in real time, in front of a customer, on a phone line you can't easily replay. The Voice Live Evaluation Harness closes that gap with four concrete capabilities: Establish a quality baseline. Run a representative audio dataset through your agent and get scores you can publish as your launch bar. Compare configurations side-by-side. Swap the underlying model (GPT-Realtime 1.5, Azure-Realtime, MAI-Transcribe-1.5), change the voice, tune VAD thresholds — and see exactly which knobs moved which scores. Catch regressions before users do. Wire it into CI and fail the build when intent resolution drops below your threshold. Optimize with data, not vibes. When task-completion drops, drill into the per-turn scores to see whether the agent failed to call the right tool, misunderstood intent, or generated an incomplete response. Keep iterating as production data rolls in. Start with the sample datasets, then grow your evaluation set with audio captured from internal testing, pilot users, and real production traffic. Re-run after every prompt tweak or model swap so the harness becomes a continuous quality signal — not a one-time launch checklist. How it works The pipeline is a five-stage loop: Audio Dataset. Multi-turn audio + expected behaviors in a simple JSONL schema. Four sample datasets ship in the repo (travel planning, complex data analytics, tool-calling tests, batch multi-conversation) so you can run end-to-end on day one. Voice Live API. Pick your Voice Live mode (Semantic VAD, PTT, or Foundry Agent), model, voice, and turn-detection settings via a JSON config file, then stream each turn of audio through the API — locally with the CLI harness, or, if you've deployed the evaluation agent, via the hosted Container App for long-running batches in your own subscription. Transcript + Response. Every turn produces an agent transcript, the model's response, and any tool calls it made — captured automatically for scoring. Foundry Evaluators. 13 built-in evaluators — powered by the same Foundry evaluator models (GPT-4.1-mini and o4-mini) used across Microsoft Foundry — judge every turn on intent resolution, task adherence, tool-call accuracy, groundedness, and more. Quality Scores. Per-turn and aggregate scores land in the Microsoft Foundry portal under your project's Evaluation tab — sortable, filterable, comparable across runs. Then loop. Audio captured from internal testing, pilots, and production traffic feeds back into the dataset — each pass makes the next evaluation more representative of what users actually do. What gets measured The accelerator ships 13 built-in evaluators out of the box, covering the dimensions that matter most for production voice agents: Category Evaluators Intent & task quality Intent Resolution · Task Adherence · Task Completion · Response Completeness Tool calling Tool Call Accuracy · Tool Call Parameter Validity · Tool Result Usage · Tool Call Success Content quality Groundedness · Relevance · Fluency · Coherence Conversational dynamics Turn-taking quality Every evaluator runs against the same Foundry evaluator models (GPT-4.1-mini and o4-mini) that power evaluation across the rest of Microsoft Foundry — so your voice-agent scores are directly comparable to your text-agent scores. Run the CLI locally against your existing Voice Live endpoint If you already have a Voice Live agent deployed and just want fast iteration on a laptop: git clone https://github.com/microsoft-foundry/voicelive-evaluation.git cd voicelive-evaluation/evaluation_harness python -m venv .venv && source .venv/bin/activate pip install -r requirements.txt cp .sample_env .env # Edit .env with your AZURE_VOICELIVE_ENDPOINT python voice_agent_evaluation.py \ --config configs/sample_vad_realtime.json The full walkthrough — dataset schema, configuration reference, score interpretation, and troubleshooting — is in the documentation. Get started Repo: microsoft-foundry/voicelive-evaluation Docs: How to evaluate Voice Live agents (preview) We'd love your feedback — try it, file issues, and tell us which evaluators you wish you had.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/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.
Now in Foundry: VibeVoice-ASR, MiniMax M2.5, Qwen3.5-9B
This week's Model Mondays edition features two models that have just arrived in Microsoft Foundry: Microsoft's VibeVoice-ASR, a unified speech-to-text model that handles 60-minute audio files in a single pass with built-in speaker diarisation and timestamps, and MiniMaxAI's MiniMax-M2.5, a frontier agentic model that leads on coding and tool-use benchmarks with performance comparable to the strongest proprietary models at a fraction of their cost; and Qwen's Qwen3.5-9B, the largest of the Qwen3.5 Small Series. All three represent a shift toward long-context, multi-step capability: VibeVoice-ASR processes up to an hour of continuous audio without chunking; MiniMax-M2.5 handles complex, multi-phase agentic tasks more efficiently than its predecessor—completing SWE-Bench Verified 37% faster than M2.1 with 20% fewer tool-use rounds; and Qwen3.5-9B brings multimodal reasoning on consumer hardware that outperforms much larger models. Models of the week VibeVoice-ASR Model Specs Parameters / size: ~8.3B Primary task: Automatic Speech Recognition with diarisation and timestamps Why it's interesting 60-minute single-pass with full speaker attribution: VibeVoice-ASR processes up to 60 minutes of continuous audio without chunk-based segmentation—yielding structured JSON output with start/end timestamps, speaker IDs, and transcribed content for each segment. This eliminates the speaker-tracking drift and semantic discontinuities that chunk-based pipelines introduce at segment boundaries. Joint ASR, diarisation, and timestamps in one model: Rather than running separate systems for transcription, speaker separation, and timing, VibeVoice-ASR produces all three outputs in a single forward pass. Users can also inject customized hot words—proper nouns, technical terms, or domain-specific phrases—to improve recognition accuracy on specialized content without fine-tuning. Multilingual with native code-switching: Supports 50+ languages with no explicit language configuration required and handles code-switching within and across utterances natively. This makes it suitable for multilingual meetings and international call center recordings without pre-routing audio by language. Benchmarks: On the Open ASR Leaderboard, VibeVoice-ASR achieves an average WER of 7.77% across 8 English datasets (RTFx 51.80), including 2.20% on LibriSpeech Clean and 2.57% on TED-LIUM. On the MLC-Challenge multi-speaker benchmark: DER 4.28%, cpWER 11.48%, tcpWER 13.02%. Try it Use case What to build Best practices Long-form, multi-speaker transcription for meetings + compliance A transcription service that ingests up to 60 minutes of audio per request and returns structured segments with speaker IDs + start/end timestamps + transcript text (ready for search, summaries, or compliance review). Keep audio un-chunked (single-pass) to preserve speaker coherence and avoid stitching drift; rely on the model’s joint ASR, diarisation, and timestamping so you don’t need separate diarisation/timestamp pipelines or postprocessing. Multilingual + domain-specific transcription (global support, technical reviews) A global transcription workflow for multilingual meetings or call center recordings that outputs “who/when/what,” and supports vocabulary injection for product names, acronyms, and technical terms. Provide customized hot words (names / technical terms) in the request to improve recognition on specialized content; don’t require explicit language configuration—VibeVoice-ASR supports 50+ languages and code-switching, so you can avoid pre-routing audio by language. Read more about the model and try out the playground Microsoft for Hugging Face Spaces to try the model for yourself. MiniMax-M2.5 Model Specs Parameters / size: ~229B (FP8, Mixture of Experts) Primary task: Text generation (agentic coding, tool use, search) Why it's interesting? Leading coding benchmark performance: Scores 80.2% on SWE-Bench Verified and 51.3% on Multi-SWE-Bench across 10+ programming languages (Go, C, C++, TypeScript, Rust, Python, Java, and others). In evaluations across different agent harnesses, M2.5 scores 79.7% on Droid and 76.1% on OpenCode—both ahead of Claude Opus 4.6 (78.9% and 75.9% respectively). The model was trained across 200,000+ real-world coding environments covering the full development lifecycle: system design, environment setup, feature iteration, code review, and testing. Expert-level search and tool use: M2.5 achieves industry-leading performance in BrowseComp, Wide Search, and Real-world Intelligent Search Evaluation (RISE), laying a solid foundation for autonomously handling complex tasks. Professional office work: Achieves a 59.0% average win rate against other mainstream models in financial modeling, Word, and PowerPoint tasks, evaluated via the GDPval-MM framework with pairwise comparison by senior domain professionals (finance, law, social sciences). M2.5 was co-developed with these professionals to incorporate domain-specific tacit knowledge—rather than general instruction-following—into the model's training. Try it Use case What to build Best practices Agentic software engineering Multi‑file code refactors, CI‑gated patch generation, long‑running coding agents working across large repositories Start prompts with a clear architecture or refactor goal. Let the model plan before editing files, keep tool calls sequential, and break large changes into staged tasks to maintain state and coherence across long workflows. Autonomous productivity agents Research assistants, web‑enabled task agents, document and spreadsheet generation workflows Be explicit about intent and expected output format. Decompose complex objectives into smaller steps (search → synthesize → generate), and leverage the model’s long‑context handling for multi‑step reasoning and document creation. With these use cases and best practices in mind, the next step is translating them into a clear, bounded prompt that gives the model a specific goal and the right tools to act. The example below shows how a product or engineering team might frame an automated code review and implementation task, so the model can reason through the work step by step and return results that map directly back to the original requirement: “You're building an automated code review and feature implementation system for a backend engineering team. Deploy MiniMax-M2.5 in Microsoft Foundry with access to your repository's file system tools and test runner. Given a GitHub issue describing a new API endpoint requirement, have the model first write a functional specification decomposing the requirement into sub-tasks, then implement the endpoint across the relevant service files, write unit tests with at least 85% coverage, and return a pull request summary explaining each code change and its relationship to the original requirement. Flag any implementation decisions that deviate from the patterns found in the existing codebase.” Qwen3.5-9B Model Specs Parameters / size: 9B Context length: 262,144 tokens natively; extensible to 1,010,000 tokens Primary task: Image-text-to-text (multimodal reasoning) Why it’s interesting High intelligence density at small sizes: Qwen 3.5 Small models show large reasoning gains relative to parameter count, with the 4B and 9B variants outperforming other sub‑10B models on public reasoning benchmarks. Long‑context by default: Support for up to 262K tokens enables long‑document analysis, codebase review, and multi‑turn workflows without chunking. Native multimodal architecture: Vision is built into the model architecture rather than added via adapters, allowing small models (0.8B, 2B) to handle image‑text tasks efficiently. Open and deployable: Apache‑2.0 licensed models designed for local, edge, or cloud deployment scenarios. Benchmarks AI Model & API Providers Analysis | Artificial Analysis Try it Use case When to use Best‑practice prompt pattern Long‑context reasoning Analyzing full PDFs, long research papers, or large code repositories where chunking would lose context Set a clear goal and scope. Ask the model to summarize key arguments, surface contradictions, or trace decisions across the entire document before producing an output. Lightweight multimodal document understanding OCR‑driven workflows using screenshots, scanned forms, or mixed image‑text inputs Ground the task in the artifact. Instruct the model to first describe what it sees, then extract structured information, then answer follow‑up questions. With these best practices in mind, Qwen 3.5-9B demonstrates how compact, multimodal models can handle complex reasoning tasks without chunking or manual orchestration. The prompt below shows how an operations analyst might use the model to analyze a full report end‑to‑end: "You are assisting an operations analyst. Review the attached PDF report and extracted tables. Identify the three largest cost drivers, explain how they changed quarter‑over‑quarter, and flag any anomalies that would require follow‑up. If information is missing, state what data would be needed." Getting started You can deploy open-source Hugging Face models directly in Microsoft Foundry by browsing the Hugging Face collection in the Foundry model catalog and deploying to managed endpoints in just a few clicks. You can also start from the Hugging Face Hub. First, select any supported model and then choose "Deploy on Microsoft Foundry", which brings you straight into Azure with secure, scalable inference already configured. Learn how to discover models and deploy them using Microsoft Foundry documentation. Follow along the Model Mondays series and access the GitHub to stay up to date on the latest Read Hugging Face on Azure docs Learn about one-click deployments from the Hugging Face Hub on Microsoft Foundry Explore models in Microsoft FoundryEvaluating AI Agents: More than just LLMs
Artificial intelligence agents are undeniably one of the hottest topics at the forefront of today’s tech landscape. As more individuals and organizations increasingly rely on AI agents to simplify their daily lives—whether through automating routine tasks, assisting with decision-making, or enhancing productivity—it's clear that intelligent agents are not just a passing trend. But with great power comes greater scrutiny--or, from our perspective, it at least deserves greater scrutiny. Despite their growing popularity, one concern that we often hear about is the following: Is my agent doing the right things in the right way? Well—it can be measured from many aspects to understand the agent’s behavior—and this is why agent evaluators come into play. Why Agent Evaluation Matters Unlike traditional LLMs, which primarily generate responses to user prompts, AI agents take action. They can search the web, schedule your meetings, generate reports, send emails, or even interact with your internal systems. A great example of this evolution is GitHub Copilot’s Agent Mode in Visual Studio Code. While the standard “Ask” or “Edit” modes are powerful in their own right, Agent Mode takes things further. It can draft and refine code, iterate on its own suggestions, detect bugs, and fix them—all from a single user request. It’s not just answering questions; it’s solving problems end-to-end. This makes them inherently more powerful—and more complex to evaluate. Here’s why agent evaluation is fundamentally different from LLM evaluation: Dimension LLM Evaluation Agent Evaluation Core Function Content (text, image/video, audio, etc.) generation Action + reasoning + execution Common Metrics Accuracy, Precision, Recall, F1 Score Tool usage accuracy, Task success rate, Intent resolution, Latency Risk Misinformation or hallucination Security breaches, wrong actions, data leakage Human-likeness Optional Often required (tone, memory, continuity) Ethical Concerns Content safety Moral alignment, fairness, privacy, security, execution transparency, preventing harmful actions Shared Evaluation Concerns Latency, Cost, Privacy, Security, Fairness, Moral alignment, etc. Take something as seemingly straightforward as latency. It’s a common metric across both LLMs and agents, often used as a key performance indicator. But once we enter the world of agentic systems, things get complicated—fast. For LLMs, latency is usually simple: measure the time from input to response. But for agents? A single task might involve multiple turns, delayed responses, or even real-world actions that are outside the model’s control. An agent might run a SQL query on a poorly performing cluster, triggering latency that’s caused by external systems—not the agent itself. And that’s not all. What does “done” even mean in an agentic context? If the agent is waiting on user input, has it finished? Or is it still "thinking"? These nuances make it tricky to draw clear latency boundaries. In short, agentic evaluations – even for common metrics like latency—are not just harder than evaluating an LLM. It’s an entirely different game. What to Measure in Agent Evaluation To assess an AI agent effectively, we must consider the following dimensions: Task Success Rate – Can the agent complete what it was asked to do? Tool Use Accuracy – Does the agent call the right tool with the correct parameters? Intent Resolution – Does it understand the user’s request correctly? Prompt Efficiency – Is the agent generating efficient and concise prompts for downstream models or tools? Safety and Alignment – Is the agent filtering harmful content, respecting privacy, and avoiding unsafe actions? Trust and Security – Do users feel confident relying on the agent? Does my agent have the right level of access to sensitive information and available actions? Response Latency and Reliability – How fast and consistent are the agent’s responses across contexts? Red-Teaming evaluations – These evaluation metrics focus on the potential misuse of agents and test for different types of attacks such as personal identifiable information leakage attacks and tool poisoning attacks. This is especially critical for non-chat completion agents — those that don’t merely chat but execute workflows, navigate APIs, or trigger automations. Their evaluation requires scenario simulation, observability instrumentation, and fine-grained analytics. Case Study: Evaluating a Simple AI Agent for Holiday Countdown To illustrate how to evaluate an AI agent, let’s walk through a simple use case: A simple AI agent for sending holiday cards. Agent Setup Platform: Microsoft Foundry Instruction: " You are an assistant agent designed to answer simple questions about the number of days remaining until holidays, and to generate and send holiday cards. Always use the tool to retrieve the current date and the official date of the holiday in question. Based on this information, calculate and provide the number of days left until the holiday. If the holiday is less than 90 days away, generate a personalized holiday card for the occasion. Then, send the holiday card as a gift to <Email address>." Tool: Bing Search, Azure Logic Apps (sending emails through Outlook) LLM Model: GPT-4.1 Example Request Query Response How many days left until New Year? "Today is July 23, 2025. New Year's Day is January 1, 2026. There are 162 days left until New Year." How many days left to Labor Day? "Today is July 25, 2025. Labor Day in 2025 falls on September 1, 2025. There are 38 days left until Labor Day! Since it's less than 90 days away, here's a personalized holiday card for the occasion:" Evaluation Dimensions Task Success Rate Goal: The agent should correctly identify the holiday and current date, then return the accurate number of days left. Evaluation: I tested 10 different holidays, and all were successfully returned. Task success rate = 10/10 = 100%. What’s even better? Microsoft Foundry provides a built-in LLM-based evaluator for task adherence that we can leverage directly: Tool Use Accuracy Goal: The agent should always use the tool to search for holidays and the current date—even if the LLM already knows the answer. It must call the correct tool (Bing Search) with appropriate parameters. Evaluation: Initially, the agent failed to call Bing Search when it already "knew" the date. After updating the instruction to explicitly say "use Bing Search" instead of “use tool”, tool usage became consistent-- clear instructions can improve tool-calling accuracy. Intent Resolution Goal: The agent must understand that the user wants a countdown to the next holiday mentioned, not a list of all holidays or historical data, and should understand when to send holiday card. Evaluation: The agent correctly interpreted the intent, returned countdowns, and sent holiday cards when conditions were met. Microsoft Foundry’s built-in evaluator confirmed this behavior. Prompt Efficiency Goal: The agent should generate minimal, effective prompts for downstream tools or models. Evaluation: Prompts were concise and effective, with no redundant or verbose phrasing. Safety and Alignment Goal: Ensure the agent does not expose sensitive calendar data or make assumptions about user preferences. Evaluation: For example, when asked: “How many days are left until my next birthday?” The agent doesn’t know who I am and doesn’t have access to my personal calendar, where I marked my birthday with a 🎂 emoji. So, the agent should not be able to answer this question accurately — and if it does, then you should be concerned. Trust and Security Goal: The agent should only access public holiday data and not require sensitive permissions. Evaluation: The agent did not request or require any sensitive permissions—this is a positive indicator of secure design. Response Latency and Reliability Goal: The agent should respond quickly and consistently across different times and locations. Evaluation: Average response time was 1.8 seconds, which is acceptable. The agent returned consistent results across 10 repeated queries. Red-Teaming Evaluations Goal: Test the agent for vulnerabilities such as: * PII Leakage: Does it accidentally reveal user-specific calendar data? * Tool Poisoning: Can it be tricked into calling a malicious or irrelevant tool? Evaluation: These risks are not relevant for this simple agent, as it only accesses public data and uses a single trusted tool. Even for a simple assistant agent that answers holiday countdown questions and sends holiday cards, its performance can and should be measured across multiple dimensions, especially since it can call tools on behalf of the user. These metrics can then be used to guide future improvements to the agent – at least for our simple holiday countdown agent, we should replace the ambiguous term “tool” with the specific term “Bing Search” to improve the accuracy and reliability of tool invocation. Key Learnings from Agent Evaluation As I continue to run evaluations on the AI agents we build, several valuable insights have emerged from real-world usage. Here are some lessons I learned: Tool Overuse: Some agents tend to over-invoke tools, which increases latency and can confuse users. Through prompt optimization, we reduced unnecessary tool calls significantly, improving responsiveness and clarity. Ambiguous User Intents: What often appears as a “bad” response is frequently caused by vague or overloaded user instructions. Incorporating intent clarification steps significantly improved user satisfaction and agent performance. Trust and Transparency: Even highly accurate agents can lose user trust if their reasoning isn’t transparent. Simple changes—like verbalizing decision logic or asking for confirmation—led to noticeable improvements in user retention. Balancing Safety and Utility: Overly strict content filters can suppress helpful outputs. We found that carefully tuning safety mechanisms is essential to maintain both protection and functionality. How Microsoft Foundry Helps Microsoft Foundry provide a robust suite of tools to support both LLM and agent evaluation: General purpose evaluators for generative AI - Microsoft Foundry | Microsoft Learn By embedding evaluation into the agent development lifecycle, we move from reactive debugging to proactive quality control.The Future of AI: The Model is Key, but the App is the Doorway
This post explores the real-world impact of GPT-5 beyond benchmark scores, focusing on how application design shapes user experience. It highlights early developer feedback, common integration challenges, and practical strategies for adapting apps to leverage the advanced capabilities of GPT-5 in Foundry Models. From prompt refinement to fine-tuning to new API controls, learn how to make the most of this powerful model.The Future of AI: Harnessing AI for E-commerce - personalized shopping agents
Explore the development of personalized shopping agents that enhance user experience by providing tailored product recommendations based on uploaded images. Leveraging Azure AI Foundry, these agents analyze images for apparel recognition and generate intelligent product recommendations, creating a seamless and intuitive shopping experience for retail customers.
