engblogs

MLflow's new tools enable developers and domain experts to efficiently build, align, and scale custom LLM judges for accurate, domain-specific evaluation of AI agents using natural language instructions and human feedback.

Leveraging Vision Language Models and EasyOCR within a Spark ML pipeline on Databricks enables scalable, cost-effective de-identification of PHI in DICOM medical images, improving privacy compliance and processing speed.

AWS and OpenAI have formed a $38 billion multi-year partnership to leverage AWS’s advanced infrastructure, including EC2 UltraServers and NVIDIA GPUs, enabling scalable and efficient deployment of OpenAI’s AI workloads such as ChatGPT and next-generation models.

IndQA is a culturally nuanced benchmark designed to evaluate AI models' understanding and reasoning in Indian languages across diverse cultural domains, addressing limitations in existing multilingual benchmarks.

MIT researchers developed FSNet, a hybrid machine learning and optimization tool that rapidly solves complex problems like power grid management while guaranteeing feasible, constraint-satisfying solutions.

AI-driven computer vision and the CODA model selection method are revolutionizing wildlife monitoring by enabling efficient, accurate analysis of vast ecological datasets to support conservation efforts.

MIT's new guidebook and podcast explore the challenges and opportunities of integrating generative AI into K-12 education, emphasizing collaboration, humility, and ongoing research to support teachers and students.

Data 360's vector search leverages GPU acceleration, optimized Kafka throughput, and AI-assisted development to enable near real-time intelligence and actionable insights on 90% of enterprise unstructured data.

Lambda and Microsoft have entered a multibillion-dollar multi-year agreement to deploy large-scale AI infrastructure powered by tens of thousands of NVIDIA GPUs, enhancing access to high-performance AI cloud computing.

Enhancements in Ray on Google Kubernetes Engine (GKE) provide a more native, efficient, and scalable experience for running AI workloads on Cloud TPUs with improved TPU-aware scheduling, simplified multi-host training, and integrated performance monitoring.

Mercado Libre leverages a multi-layered Spanner architecture, integrated with the Fury platform and advanced analytics, to achieve global consistency, scalability, and agility while enabling AI-driven applications and operational excellence.

Discover how Gemini Enterprise, Code Assist, and CLI integrate AI agents with Google Cloud's HPC to accelerate and enhance scientific research workflows from hypothesis generation to publication.