engblogs

Crypto.com enhances its enterprise AI assistants' efficiency and accuracy by leveraging large language model reasoning and iterative feedback-driven prompt optimization within a modular architecture on AWS.

Explore how the Amazon Nova Act SDK enables fast, reliable, and secure production-ready browser automation agents using agentic AI and AWS integrations for scalable enterprise workflows.

Leverage Strands Agents SDK and Amazon Bedrock foundation models to build an AI-powered drug discovery research assistant that integrates multiple scientific databases and generates comprehensive reports.

Lotus’s leverages AI-powered Natural Language Query and chatbot technology integrated with Databricks infrastructure to provide 3,000+ stores with real-time, personalized retail intelligence that empowers faster, smarter decision-making.

Explore Google Cloud Security innovations, AI-driven threat detection, Mandiant expertise, and hands-on training at Black Hat USA 2025 to enhance your organization's cybersecurity posture.

The White House AI Action Plan outlines a comprehensive strategy focused on accelerating AI innovation, building secure American AI infrastructure, and leading international AI diplomacy, with strong emphasis on open-source models, cybersecurity, balanced export controls, and public-private collaboration to maintain U.S. leadership in AI.

mRAKL leverages retrieval-augmented generation to enhance multilingual knowledge graph construction for low-resourced languages by reformulating it as a question answering task and employing cross-lingual transfer.

MMAU benchmark offers a comprehensive and reproducible evaluation framework for assessing large language models' capabilities across diverse domains and core skills without complex environment setups.

Exploring how integrating global calibration into multiaccuracy significantly enhances its effectiveness in achieving stronger fairness guarantees and agnostic learning capabilities.

Investigating how augmenting LLM-based annotation systems with external validation tools like web-search and code execution can enhance the quality of pairwise preference data in complex domains such as long-form factual content, math, and code.

ASPERA introduces a simulated environment and dataset to rigorously evaluate large language models' ability to generate complex action execution plans using custom assistant libraries.

Investigating how targeted neuron interventions enhance cross-lingual alignment and improve retrieval accuracy in multilingual language models without costly fine-tuning.