How can we ensure that AI-generated code maintains high quality, adheres to good style, and follows best practices? This question has been on my mind recently, and it is the topic of my latest blog post: ➡️

In the third and last part of my RAG evaluation blog series, I show how to improve a RAG pipeline on @googlecloud's VertexAI using @confident_ai DeepEval's RAG triad metrics: ➡️

In my latest step-by-step guide, I show how to evaluate a RAG pipeline with RAG triad metrics using @confident_ai's open source LLM framework DeepEval and @googlecloud's Vertex AI:

In my latest blog post, I talk about how to evaluate RAG pipelines, what methods and metrics to use, and explore the RAG triad:

Is it just me or @youtubemusic got much better recently? I've been a long time Spotify user but I recently found YouTubeMusic better in exploring new music

In this short & sweet blog post, I show how easy it is to switch between Gemini 2.0 on Google AI and Vertex AI with the new unified Google Gen AI SDK:

@airasia Your 'Chat with Bo' AI bot is useless. I've been trying to cancel a flight and it's been impossible to talk to anyone in the chat, other than the useless AI bot, and when I call on the phone, it just directs me to a voice message. Bad customer service.

✨ Ready to dive into Gemini 2.0 Flash on Vertex AI? Check out this essential guide from my colleague Erwin Huizenga: 🚀 #GenAI #Gemini2

ℹ️ Tired of chaotic outputs from language models❓In my newest blog post, I explain how to define response schemas and parse LLM outputs to Python classes with LangChain and Pydantic:

@ivanburazin Thankfully we have great public transport and we never have to drive to/from central London

LLMs feel like a new frontier, but they also face old challenges: like figuring out which calls are slow or failing. Observability is key—logging, tracing, metrics. Learn how to trace LLM calls with OpenTelemetry & @langtrace_ai in this post. 👇

Avoid common LLM pitfalls: Lessons learned building a GenAI powered app:

RT @glaforge: #OWASP top 10 for #LLM applications 1⃣prompt injection 2⃣sensitive information disclosure 3⃣supply chain 4⃣data and model po…

@DavidGiard Thanks!

Batch generation in Gemini lets you process multiple generative AI requests in parallel and 50% cheaper! Perfect for scaling tasks like creating descriptions for thousands of books in one go. ➡️ Check out the details in my latest blog post:

I’ve been exploring LLM evaluation and security frameworks lately because ensuring LLMs behave properly is a tough problem. In this week’s post, I explore LLM Guard and Vertex AI: ➡️ #GenerativeAI #LLM #evaluation #security @GoogleCloudTech

In my latest blog post, I explore Promptfoo, an open-source testing framework, and how it can be used with Vertex AI for robust evaluation and security testing of your LLM apps: #LLM #Promptfoo #VertexAI #GoogleCloud

If you have an M1 Mac, you might run into issues with Homebrew installed packages as the default location for the packages changed in M1. Thankfully, @joshalletto's blog post explains the details:

🚀 In the third and last post of the series, I show how to store and use image embeddings with #Firestore and #LangChain ➡️ #AI #LLM #GenAI #GoogleCloud #VertexAI