Amazon Releases Jev-Inspired Decision Model Amid Surge in AI Agents

Amazon Web Services (AWS) has released Strands Decider 2B, an open-source decision model drawing inspiration from TypeSafe's Jev. The model is designed to cater to developers seeking AI intelligence better suited for computer automation than frontier large language models (LLMs).

Strands Decider 2B functions as a high-speed, low-cost method for sorting through pre-decided options, providing a confidence score for its selections. The model is fully open-sourced, available immediately, and small enough to run locally. AWS's release comes in the same week that OpenAI announced a similar offering.

The project originated from Amazon distinguished engineer Marc Brooker, who was inspired by Jev and sought to build his own iteration of such a model. Brooker's initial home-built project proved successful, prompting AWS engineers to refine and release it through Strands Labs, an organization focused on developing tools and protocols for deploying AI agents.

Brooker explained that the need for this type of tool arose from discussions with AWS customers whose agentic workflows did not always necessitate the capability or expense of a full-fledged LLM. Strands Decider offers customers a workflow step that is structured to be more reliable due to confidence scores and a closed domain of answers, while also offering lower latency and potentially lower costs.

Similar to other decision models, Strands Decider is built upon the "torso" of an LLM, in this case Qen3.5-2B. Instead of generating text, it outputs a predefined set of options along with a confidence score for its chosen option.