3️⃣ 2️⃣ 1️⃣ Our free course on LLM evaluations for AI product teams starts today! 🎥 7 days of byte-sized videos into your inbox ⭐️ Certificate upon completion 👩‍💻 No coding skills required 👩‍🎓500+ students have signed up You can still join the course👇 https://t.co/Go2bNYJXCR

📌 In case you missed it ⚖️ How to build LLM judges that mirror human judgment? Our tutorial shows how to design, test, and calibrate an evaluator for evaluating the quality of LLM-generated code reviews. You can adapt the example for your use case: https://t.co/G1vVvBrXeg

💬 Writing good LLM prompts is tedious. Join us for a webinar on automated prompt optimization: 🐣 Intro to prompt optimization ✅ How to improve prompts with Evidently Open-source 💻 Live demo Thanks @DataTalksClub for the invite! Register here 👇 https://t.co/ajSePLx59i

📌 In case you missed it 7 questions about LLM judges! We answered some of the most common questions we get about how LLM judges work and how to use them effectively 👇 https://t.co/LnCc8TyCDZ

A Friday ML use case 📕 📚 From the database of 650 ML & LLM systems: https://t.co/jJoUj6MfFZ How Booking built an AI agent that assists partners by automatically suggesting a relevant response to each guest inquiry. https://t.co/isWy5be7iu

10 AI risks every team should be testing for! ⚠️ Top risks when building AI products ❌ How to mitigate them ✍️ Common AI risk assessment frameworks ✅ How to set up a continuous AI testing workflow Check whether you test for all the critical risks: https://t.co/geKkIALIqq

🕶 Detecting and reducing scheming in AI models AI models can “scheme” – act helpful while secretly pursuing other goals. Research from OpenAI and Apollo suggests “deliberate alignment” as a new approach to reducing covert actions. Read the paper 👇 https://t.co/qQleAMICJ5

📌 In case you missed it 7 agentic AI examples and use cases We explore seven agentic AI examples and use cases in the real world – from transaction analysis to e-commerce recommendations to code review 👇 https://t.co/3J48HOhKYd

A Friday ML use case 📕 📚 From the database of 650 ML & LLM systems: https://t.co/jJoUj6MfFZ How LinkedIn engineered its Hiring Assistant: the architecture, design choices, and lessons learned from building an agentic product. https://t.co/hnA8O4SNrM

🚀 LLM tracing and dataset management are now live in Evidently open-source! The new release unlocks previously closed features: 🔡 Data storage backend ✅ Raw dataset management and viewer 🐾 LLM tracing storage and viewer Try it now 👇 https://t.co/xMnU1xHb7O

☠️ How LLMs could be insider threats Anthropic stress-tested 16 LLMs and identified potentially risky agentic behaviors. Turns out, LLMs can blackmail and leak sensitive info as a means to avoid replacement or achieve their goals. Read the paper 👇 https://t.co/JeLFbbmqSP

📌 In case you missed it Gen AI use cases in 2025: learnings from 650 examples. We highlighted some new patterns of how top companies apply Gen AI based on a database of real-world AI and ML use cases we’ve been curating 👇 https://t.co/WfL237WONA

A Friday ML use case 📕 📚 From the database of 650 ML & LLM systems: https://t.co/jJoUj6MfFZ How Thomson Reuters uses RAG to enhance customer service: support execs can retrieve relevant info from a curated database through a conversational interface. https://t.co/vnmboQoZzz

⚖️ How to align LLM judge with human labels: new tutorial! We break down the process of designing, testing, and calibrating LLM evaluators. We also show how to create an LLM judge for evaluating the quality of LLM-generated code reviews👇 https://t.co/G1vVvBrXeg

🦄 Why language models hallucinate OpenAI’s research suggests that standard training and evaluation reward guessing over acknowledging uncertainty. Read the paper: https://t.co/9znwcKbcL6

📌 In case you missed it 25 AI benchmarks: examples of AI models evaluation A brief explainer of what AI benchmarks are and why we need them, with 25 examples of common AI benchmarks for reasoning, conversation abilities, coding, RAG, and tool use 👇 https://t.co/xPonNn2CN4

A Friday ML use case 📕 📚 From the database of 650 ML & LLM systems: https://t.co/jJoUj6MfFZ How Grab combines vector similarity search with LLMs to enhance the relevance and accuracy of search results for complex queries. https://t.co/FBPXG5cLNw

❓ 7 questions about LLM judges! We answered some of the most common questions we get about how LLM judges work and how to use them effectively 👇 https://t.co/LnCc8TyCDZ

📌 In case you missed it 8 AI hallucinations examples 🦄 We put together eight examples of real-world AI hallucinations – from a transcription tool fabricating texts to citing made-up company policies. Explore the examples 👇 https://t.co/Mz8QQhoy7C

📊 What's your go-to MLOps stack? The State of MLOps Survey is live – and we’re excited to see the first results, where Evidently is the most popular tool for ML monitoring 🔥 Kudos to @AxSaucedo for this insightful research! You can still vote here 👇 https://t.co/Rsv1dMgDnd

A Friday ML use case 📕 📚 From the database of 650 ML & LLM systems: https://t.co/jJoUj6MfFZ How Instacart helps users find new products by incorporating LLMs into the search stack to generate discovery-oriented content. https://t.co/kFtDrOtHzu