Apporto gains data-driven clarity to scale AI grading with confidence
A six-model evaluation on AWS identified a path to cut grading costs by up to 75% while preserving academic integrity and student mastery.
Industry
Technology
Challenge
Apporto needed an unbiased way to evaluate large language models to balance cost, quality and scalability before scaling its AI-powered tutoring, grading and academic integrity tools for higher education.
Outcome Apporto now has a data-driven understanding of six LLMs across cost, accuracy and scalability, identifying a potential path to reduce grading costs from about $20 per paper to about $5 per paper. Solutions AI Platforms Amazon Web Services (AWS)Customer
Apporto is an employee-owned software company that supports digital learning with flexible virtual computing and AI tools for education, supporting universities like DeVry and ASU. The platform is designed to run seamlessly across cloud, on-premises and hybrid environments. The company built its foundation in virtual desktop access for universities, enabling students to use academic applications from virtually any device through a browser-based experience.
Today, Apporto is advancing this foundation through its AI Product Suite, a platform designed to support the full learning lifecycle while maintaining academic rigor. The suite is built on four pillars: CoTutor, which guides student reasoning; PowerGrader, which delivers rubric-aligned feedback; TrustEd, which provides evidence-based insight into how work is created; and ExamSpace, which enables secure, browser-based testing environments.
This approach is grounded in the concept of cognitive friction, helping ensure that AI supports learning without removing the mental effort required for mastery. Rather than optimizing for speed alone, Apporto focuses on making the learning process observable and defensible for institutions and faculty.
By focusing on cognitive friction, Apporto helps AI support deep engagement rather than cognitive offloading. As the company prepares to scale these capabilities, it must now identify the optimal large language models to balance quality, consistency, and cost.
“The relationship between Rackspace and Apporto was key in helping us understand the AI landscape and the capabilities of various LLMs. Their broad knowledge led us to new avenues to provide the best results to students and faculty.”
Situation
Apporto recognized the rapid adoption of AI in higher education, along with the growing risk that these tools could undermine the learning process. While much of the industry focused on detecting AI-generated content, Apporto identified a more fundamental issue: when student work is produced without visible reasoning, traditional signals of mastery begin to break down. Faculty are left evaluating outputs that appear correct but may not reflect genuine understanding.
To address this, Apporto built its strategy around cognitive friction, helping ensure that AI supports learning while preserving the effort required for students to develop true competence. This required more than product design. It demanded that every output, especially in grading workflows, remain consistent, explainable and aligned with academic standards. As PowerGrader adoption expanded, maintaining that balance at scale became increasingly complex.
Selecting the right large language models was critical to achieving this goal. Apporto needed to evaluate models not only for response quality, but for their ability to perform reliably under real grading conditions, including bulk submission scenarios and rubric-based assessment. At the same time, cost efficiency was essential to ensure long-term scalability for institutional use.
Rather than relying on internal testing or vendor claims, Apporto sought an objective, data-driven evaluation. It engaged Rackspace Technology, with support from the AWS Generative AI Innovation Center, to assess multiple leading models against its actual grading workflows and performance requirements.
“Rackspace provided the strategic insights we needed to scale our AI suite effectively. Our university customers need tutoring, grading and academic integrity tools delivered at a reasonable cost, and Rackspace helped us better understand how to achieve that.”
Solution
Rackspace Technology worked with Apporto to design and execute a structured evaluation of large language models aligned to the real demands of its PowerGrader solution. Rather than relying on synthetic benchmarks, the engagement focused on Apporto’s actual grading workflows, prompts and rubric-based assessment criteria to help the results reflect real academic use cases.
Using Amazon Bedrock as a controlled environment, Rackspace evaluated six leading models across key dimensions including response quality, consistency, latency and cost. The team simulated bulk submission scenarios to mirror peak academic workloads, where large volumes of assignments are submitted simultaneously. This approach allowed Apporto to understand how each model would perform under realistic operating conditions, not just in isolated tests.
Rackspace also applied prompt engineering and testing methodologies to assess how well each model could produce structured, rubric-aligned feedback. This was critical for maintaining consistency and helping ensure outputs remain explainable and defensible for faculty. By comparing performance across multiple models in parallel, Apporto gained clear visibility into tradeoffs between cost, accuracy and scalability.
Rackspace provided Apporto with an objective, data-driven foundation for model selection. Instead of committing to a single vendor based on assumptions, Apporto can now make informed decisions based on performance within its own workflows, while retaining the flexibility to adapt as models evolve.
Rackspace helped Apporto identify a path to reduce AI grading costs by up to 75%.
Outcome
Through this engagement, Apporto gained the clarity and confidence needed to scale its AI capabilities with precision. By evaluating six large language models against its own grading workflows, the company established a data-driven understanding of how model selection impacts cost, performance, and academic quality in real-world conditions.
Rackspace identified a path to reduce grading costs from approximately $20 per paper to about $5 per paper, while maintaining the high standards required for higher education. This outcome demonstrates that cost efficiency and academic rigor do not have to be competing priorities when models are selected and validated against the right criteria.
Apporto now has the ability to compare models based on consistent benchmarks, enabling more informed decisions as new models emerge and performance evolves. This flexibility allows the company to adapt its strategy over time without disrupting its platform or compromising the integrity of student outcomes.
The engagement also reinforced Apporto’s commitment to cognitive friction, ensuring that AI remains a tool for supporting learning rather than bypassing it. With a validated model strategy in place, Apporto is positioned to expand its AI Product Suite with greater confidence, delivering scalable, reliable solutions that strengthen both learning outcomes and institutional trust.
About Rackspace Technology
Rackspace Technology is the multicloud solutions expert. We combine our expertise with the world’s leading technologies — across applications, data and security — to deliver end-to-end solutions. We have a proven record of advising customers based on their business challenges, designing solutions that scale, building and managing those solutions, and optimising returns into the future.
As a global, multicloud technology services pioneer, we deliver innovative cloud capabilities to help customers build new revenue streams, increase efficiency and create incredible experiences. Recognised as a best place to work, year after year, by Fortune, Forbes, Great Places to Work and Glassdoor, we attract and develop world-class talent to deliver the best expertise to our customers. Everything we do is underpinned by an obsession with our customers’ success — our Fanatical Experience™ — so they can work faster, smarter and stay ahead of what’s next.
Let’s Talk Strategy
Tell us a little about your challenges and we’ll contact you.
You may withdraw your consent to receive additional information from Rackspace Technology at any time. Information collected in this form is subject to the Rackspace Technology Privacy Notice.
Rackspace Technology Support
To create a ticket or chat with a specialist regarding your account, log into your account.
Support Phone
0800 988 0300
International Support
+1-512-361-4935
Help Documentation
System status
Rackspace Technology Careers
Rackspace Technology accelerates the value of the cloud during every phase of a customer’s digital transformation. Join us on our mission.