engblogs

Optimizing LLM inference acceleration on NVIDIA GPUs with ReDrafter

Explore how Česká spořitelna implemented GenAI in call centers to enhance quality control and reduce operational costs while managing approximately 2 million calls annually.

Improving iOS performance in the Threads app by monitoring key metrics, including %FIRE, TTNC, and cPSR, while tackling performance challenges and case studies on publish reliability and navigation latency.

Integrating Amazon Q Business with ServiceNow for enhanced user productivity and insights through generative AI-powered assistants and custom plugins

Fastweb fine-tuned Mistral model using SageMaker HyperPod as first step to build Italian large language model

Exploring conversations shaping Duolingo Design through Duologues speaker series

FACTS Grounding introduces a new benchmark and leaderboard to rigorously evaluate and improve the factual accuracy and grounding of large language model responses using comprehensive, multi-judge assessment across diverse, real-world documents.

Enhancing large language model (LLM) evaluation using Snorkel Flow for better models and faster results

Manage queries in version control with Git support in Databricks' New SQL Editor.

Exploring the latest features of OpenAI o1 and innovative developer tools

Evaluating models performance on Domain Intelligence Benchmark Suite (DIBS) in comparison to academic benchmarks to highlight the need for domain-specific testing and choosing models based on specific needs.

Empowering successful title launches and discoverability at Netflix through robust observability solutions.