At Moveworks, each new foundation model release is an opportunity to elevate the ability of our AI Assistant to drive more value across your enterprise.
That’s why we’re excited to announce that the Moveworks Reasoning Engine is now powered by GPT-5.1 — OpenAI’s most advanced flagship model yet. This upgrade brings major gains in performance, accuracy, and transparency, enabling our AI Assistant to unlock faster, more insightful support across your organization.
Our commitment to best-in-class technology
To deliver the best outcomes for our customers, our guiding principle is to stay technology-agnostic. We continually evaluate models from OpenAI and other research labs to adopt the best technology for each job.
Moveworks Reasoning Engine is a sophisticated hybrid AI system that autonomously understands, plans, executes, and adapts to employee requests by orchestrating multiple models from different providers, each optimized for specialized tasks.
GPT-5.1 now powers several of the highest-impact reasoning and planning functions in this system – including plugin orchestration, task decomposition, and knowledge synthesis – while other models from Meta, Google, Snowflake, and other providers handle domain-specific workloads. This enables each task is handled by the model best suited for it.
Because we maintain this hybrid AI infrastructure for you, your Assistant just works. Moveworks handles security testing, performance evaluations, and model upgrades behind the scenes so you don’t have to. This allows you to stay focused on rollout and adoption, while your Assistant gets smarter and more capable over time.
Our core directive is to make employee support radically more efficient, and we’re excited that GPT-5.1 is the next step in that journey.
Key benefits of GPT-5.1 in Moveworks, rigorously validated
Before adopting OpenAI’s GPT-5 family of models, we spent months running rigorous evaluations, including:
- Side-by-side model tests - Validating improvements on real, current use cases by comparing model versions.
- Golden set checks - Measuring accuracy on core, pre-defined plugin tasks (trigger, argument filling).
- Regression analysis - Confirming that new changes haven't negatively impacted performance on previously strong, established capabilities.
- Human review on live traffic - Assessesing overall user experience quality and nuanced differences (summary, citations).
Because our AI Assistant is entrusted to take action and update live business systems for millions of employees around the world, we only upgrade when the new model shows proven, measurable improvements for our customers.
Our evaluations show that GPT-5.1 delivers clear, measurable gains on user experience. It also offers better latency relative to GPT-5, letting us tap into its higher reasoning capabilities without affecting response times.
- Smarter, more relevant tool use: We see that it’s able to find the right tools by executing more in parallel, solving requests faster and more efficiently. Reduction in SmartHandoffPlugin calls indicate that the model is more likely to resolve requests independently before escalating to other systems or human help.
- More relevant search summarization capabilities: We are able to look across a larger set of relevant sources and summarize the information in a way that is useful to the user. One outcome is higher information intensity in the returned responses – they are 14% longer responses, with 39% more resources cited.
- Personalized search results: Another improvement is that the AI Assistant is able to better tailor answers to each employee based on location, country code, role, department, and other custom attributes that matter to your organization. By bringing these attributes into working memory, the Assistant gives a more personalized search experience.
- New experiences unlocked: GPT-5.1 lays the groundwork for future improvements, like richer reasoning steps to help you understand how the Assistant thinks, response streaming, and the ability to choose its extended thinking levels. Stay tuned for these updates coming soon.
Live today, no action required
We’re excited to note that the Moveworks Reasoning Engine is now powered by GPT-5.1. No setup work or downtime required on your end — your AI Assistant has already been upgraded.
(Note: Few account exceptions apply, please contact your CSM for more information)
Alongside our Reasoning Engine, Quick GPT – Moveworks’ enterprise-secure generative AI feature for research and content creation – also upgraded to GPT-5 mini’s enhanced reasoning and clarity, producing more structured, context-aware content with relevant citations for research and writing.
Moving forward
Moveworks continues its commitment to innovation without compromise, delivering measurable improvements and a new level of transparency. We continue to be excited for what’s next, with every step of foundation model evolution.
If you’re interested in exploring these advancements for your workforce, connect with your Moveworks account team or reach out to learn more.
Table of contents
