LLAMA 3.1 8B

Litmos

Litmos image
Aitomatic logo
CASE STUDY

Improving search and engagement with Llama-powered learning assistant

At a glance

Industry: Technology and software

Use case: Powering natural language search with AI for faster content discovery

Goal: Increasing engagement, outcomes and adoption of immersive learning

Llama versions: Llama 3.1 8B

Deployment: Self-hosted on Kubernetes, Azure AKS

95%
reduction in time to complete content authoring tasks
140%
improvement in MMLU-PRO benchmark accuracy
59%
improvement in MMLU versus previous models
110%
increase in IFEVAL benchmark performance

*All results are self-reported and not identifiably repeatable. Generally, expected individual results will differ.

THEIR STORY

Innovative learning solutions fueling business success

Litmos develops learning solutions for over 4,000 top-performing companies and 30 million users. Its learning solutions include the easiest-to-use learning management solution (LMS), comprehensive learning content collections, integrations with top workflow tools and services to support success.

THEIR GOAL

Replace keyword-based search with an intuitive, AI-powered experience

The Litmos LMS has vast libraries of learning materials in text, images and video in unique file formats. However, the platform’s keyword-based search made it difficult to find content, a frustrating experience that discouraged learning and lowered engagement. The Litmos team wanted to bypass the constraints of exact keyword matches with an AI-powered assistant that could help learners search through natural conversations.

THEIR SOLUTION

Llama-powered AI transforms content discovery with natural language

Litmos replaced its keyword-based search with a Llama 3.1 8B-powered assistant, allowing learners to use natural, conversational language to find courses and materials. For example, a learner might ask, “What are the best courses to improve my project management skills?” The assistant summarizes completed courses and asks follow-up questions to reinforce learning.

THEIR APPROACH

Portable Llama improves performance and security

Litmos chose Llama 3.1 8B for its combination of cutting-edge performance, small footprint and portable, open-source licensing. The assistant combines off-the-shelf Llama and a retrieval augmented generation (RAG) workflow with prompt engineering in a Kubernetes deployment running on private, Azure AKS GPU clusters. By eliminating API calls to external AI providers, the solution safeguards proprietary content and learner data while meeting Litmos’s clients’ enterprise security requirements.

their solution graphic

The Llama-powered Litmos assistant transformed keyword search into a natural language discovery and learning tool.

THEIR SUCCESS

Llama-powered assistant improves learning, fuels platform adoption

Litmos’s Llama-powered assistant has transformed how learners discover content and engage with the Litmos LMS. By eliminating the friction of traditional keyword-based search, the platform encourages deeper interaction, making learning more seamless and engaging. Improving the learning experiences is expanding adoption of the platform and helping Litmos solidify its position as an innovator in AI-driven eLearning solutions.

Projected results:
• 95% reduction in time to complete content authoring tasks
• 140% improvement in MMLU-PRO benchmark accuracy
• 59% improvement in MMLU versus previous models
• 110% increase in IFEVAL benchmark performance
*All results are self-reported and not identifiably repeatable. Generally, expected individual results will differ.
The AI assistant, powered by Llama, is a game-changer. With it, we’ve set a new standard in eLearning solutions. Learners engage more, administrators work smarter and the entire experience feels seamless.
“The AI assistant, powered by Llama, is a game-changer. With it, we’ve set a new standard in eLearning solutions. Learners engage more, administrators work smarter and the entire experience feels seamless.”

Tommy Richardson, Chief Product and Technology Officer, Litmos

Models used

Create generative AI applications for business with open-source large language models that bring unmatched control, customization and flexibility.
Start building
applications icon
Text

Llama 3.1 8B

Optimized for multilingual dialogue use cases
Outputs text and code, 128k context length
*Licensed under Llama 3.1 Community License Agreement
Download models
Horizon banner image

Stay up-to-date

Our latest updates delivered to your inbox

Subscribe to our newsletter to keep up with the latest Llama updates, releases and more.

Explore more
Dec 18, 2025
Tinfoil
Keeping data and models private, even when running in the cloud
Tech
Read story
Dec 10, 2025
Stoque
Unifying internal knowledge for faster insights and efficiency
Tech
Read story
Nov 14, 2025
Contextual AI
Reducing hallucinations with a better context layer for AI agents
Tech
Read story