Score Labs Emerges From Stealth on the Intel Stage at Microsoft Ignite 2026, Showing Enterprise AI at a Fraction of the Cost on Intel Xeon

Score Labs Emerges From Stealth on the Intel Stage at Microsoft Ignite 2026, Showing Enterprise AI at a Fraction of the Cost on Intel Xeon

Score Engine runs each AI workload on the right hardware across CPUs, GPUs and hybrid systems, cutting inference costs while keeping sensitive data under company control

Score Labs, an enterprise AI inference company, today emerged from stealth and announced it will present on the Intel Booth Theater stage at Microsoft Ignite 2026. In a session on Thursday, Nov. 19, at 10:00 a.m. PT, Score Labs will show four enterprise AI use cases running live on Score Engine, the company's inference software, on standard Intel Xeon servers. Score Labs will also run hands-on demos at the Intel booth twice daily throughout the conference.

Enterprise AI has moved past short chatbot prompts. Companies now want AI to read hundreds of pages of contracts, work through patient records, review transactions, and write and test code, often with many people using it at once. That kind of work is limited less by raw processing power and more by memory. It is also exactly where the cost of GPU-only infrastructure, hardware shortages and power limits are forcing companies to scale back their AI plans. And much of this work involves data that companies cannot or will not send to an outside AI service.

Standard x86 servers, like the Intel Xeon systems already running in most data centers, have the memory this work needs. What has held them back is the lack of inference software built to get full performance out of them. Score Engine closes that gap. Score Labs designed its inference stack from the ground up to run top open-weight AI models on x86 servers with fast, reliable performance, delivering 30% to 4x more performance than widely used open-source serving software such as vLLM and SGLang. Score Engine is tuned to the math and memory of each chip it runs on, so it can send each AI job to the hardware best suited for it. Companies cut what they spend on AI and keep sensitive data on infrastructure they own and control.

"For two years the industry has assumed every AI job needs the most expensive chip you can buy," said Sean Cook, CEO of Score Labs. "For the work enterprises actually do – long documents, AI agents, private data – that assumption is costing them a fortune and slowing them down. Score Engine lets companies run on the right hardware for the job, from Intel Xeon servers to GPUs to hybrid systems, under their control, at a fraction of the cost. Microsoft Ignite is where we start showing people what that looks like."

Example use cases, live on Intel Xeon:

  • Legal due diligence: AI reads an entire deal data room, flags risks and produces a finished presentation for the deal team, without documents ever leaving the company's environment.
  • Anti-money laundering review: AI works through large volumes of transaction records and case files to surface suspicious activity for human investigators, with sensitive financial data kept in-house.
  • Freight and logistics operations: AI processes load and shipping documents, checks them against company policy and routes exceptions to a person for approval.
  • AI coding agents: A single server supports 8 to 16 coding sessions running at the same time, giving engineering teams AI assistants without sending source code to an outside service.
  • Multimedia AI studio: A single server handles several multimodal requests at once, such as text-to-speech, speech-to-text, video summarization, image generation and decision-making.

Built for proof, not promises

Score Engine is already running in production. In one example, a dental AI platform replaced a commercial AI service with Score Engine. It cut its projected monthly inference cost by more than 80% and got better results than its previous approach.

"Score Labs showed us that we did not have to trade quality for control or cost," said the Founder and CEO of the dental AI platform. "We achieved better results than our previous model approach while reducing our projected inference expense by more than 80%. Just as important, we now have a path to keep sensitive information within the infrastructure we control."

In testing validated by Intel on long-running AI agent tasks, Score Engine running on servers with four Intel Xeon 6 or 5th Gen Intel Xeon processors delivered 1.5x better cost efficiency than an NVIDIA H100 GPU. It also ran the same work at 3x to 17x lower cost than hosted model APIs accessed through OpenRouter.

Engineered from the ground up

Rather than layering on top of existing open-source serving tools, Score Labs built its own low-level software stack from scratch: six proprietary libraries that control how AI models use a processor's math units, memory and network connections. The company has filed provisional patents covering its technology.

See Score Labs at Microsoft Ignite

  • Intel Booth Theater session: Thursday, Nov. 19, 10:00 a.m. PT, presented by Sean Cook, CEO, and Ashish Kumar, Co-founder and Chief AI Systems Architect
  • Live demos: Intel Booth, two one-hour sessions daily Nov. 17-20
  • Contact: scorelabs.ai/contact or contact@scorelabs.ai

About Score Labs

Score Labs builds enterprise AI inference software that runs every AI workload on the right hardware for the job. Its Score Engine lets companies run leading open-weight AI models on CPUs, GPUs and edge devices, including existing infrastructure and new deployments, from the cloud to on-premises data centers to the edge. Customers get lower cost, consistent and repeatable results, and full control over their data. Score Labs serves enterprises in healthcare, legal, financial services and software development, and works with technology partners including Intel and Microsoft. Learn more at scorelabs.ai.

Intel, the Intel logo and Xeon are trademarks of Intel Corporation or its subsidiaries. Microsoft and Azure are trademarks of the Microsoft group of companies. NVIDIA and H100 are trademarks of NVIDIA Corporation. All other names are the property of their respective owners.

Media contact: media@scorelabs.ai | www.scorelabs.ai

View source version on businesswire.com: https://www.businesswire.com/news/home/20261117452751/en/

The articles, information, and content displayed on this webpage may include materials prepared and provided by third parties. Such third-party content is offered for informational purposes only and is not endorsed, reviewed, or verified by Morningstar.

Morningstar makes no representations or warranties regarding the accuracy, completeness, timeliness, or reliability of any third-party content displayed on this site. The views and opinions expressed in third-party content are those of the respective authors and do not necessarily reflect the views of Morningstar, its affiliates, or employees.

Morningstar is not responsible for any errors, omissions, or delays in this content, nor for any actions taken in reliance thereon. Users are advised to exercise their own judgment and seek independent financial advice before making any decisions based on such content. The third-party providers of this content are not affiliated with Morningstar, and their inclusion on this site does not imply any form of partnership, agency, or endorsement.

Popular

Sponsor Center