Reactive Machines

A financial services team spends significant time each week manually creating client PowerPoint presentations from quarterly reports, compliance documents, and…

Discrete diffusion models have emerged as strong alternatives to autoregressive language models, with recent work initializing and finetuning a base…

Code-switching (CS), the alternation between languages within an utterance, is common in multilingual speech but underrepresented in pretraining corpora. Increasing…

AI agents increasingly need to access web resources that are priced per request. For example, premium datasets charge by the…

Managing AI model access across dozens or hundreds of AWS accounts creates operational friction when each workload account must subscribe…

Security analysts need fast answers about their cloud infrastructure without deep familiarity with domain-specific query languages. Orca Security’s tool covers…

Mixture-of-Experts (MoE) is a sparsely-activated architecture that scales large language model (LLM) capacity while keeping per-token compute low. Because its…

For most languages of the world, language model pre-training operates in a data-constrained regime where models must repeat their training…

Single-Region inference can throttle when demand exceeds local capacity. Teams in Thailand, Malaysia, Singapore, Indonesia, and Taiwan need access to…

Applications that route requests, select agent actions, or assign policy outcomes often need scores over a fixed set of choices…