Veterinary medicine is a demanding field where every decision carries weight. Vets care for many species. They often work alone and under pressure, while balancing growing demand with limited resources and minimal insurance support. These constraints make it increasingly difficult to deliver the consistent, high-quality care patients deserve.Recognizing these realities, AITEM Solutions, a member of the NVIDIA Inception program for startups, builds AI systems that meaningfully extend human capability. By applying advanced digital healthcare technologies to veterinary medicine, the company aims to enhance precision, efficiency and clinical decision-making in complex diagnostic environments.
THEIR GOAL
Improving clinical decision-making and reducing mental load
The AITEM Solutions team set out to develop a domain-specific AI assistant tailored to the complexity of veterinary medicine. The goal was to help vets manage daily clinical tasks under time pressure, handle the variability of cases and compensate for limited access to specialists. Ultimately, they sought to improve the speed and quality of clinical decision-making while easing the cognitive and emotional demands placed on veterinarians.
THEIR SOLUTION
A domain-specific veterinary AI assistant backed by Llama
AITEM Solutions built LAIKA, the first AI assistant designed specifically for veterinary care. Initially developed with Llama 2 and later upgraded to Llama 3.1 70B Instruct, LAIKA acts as a knowledgeable colleague — providing insights for complex cases or a second opinion to confirm diagnostic reasoning. It can also support client communication, helping vets navigate sensitive conversations with pet owners.Integrating Llama into the assistant enabled LAIKA to deliver fast, accurate responses during everyday clinical workflows. Through a specialized retrieval-augmented generation (RAG) pipeline that incorporates the NVIDIA NeMo Retriever reranking model via NVIDIA Inference Microservices (NIM), LAIKA grounds its outputs in relevant medical literature and patient-specific data to reduce diagnostic errors.
THEIR APPROACH
Llama stands out with its openness and flexibility
AITEM Solutions selected Llama as LAIKA’s foundation model for its strong balance of response quality, inference speed and openness. Its open weights enabled fine-tuning for veterinary use, while flexible deployment options — raw model, NVIDIA NIM microservice, or serverless API — helped avoid vendor lock-in.Using NVIDIA NIM, the team deployed the NeMo Retriever reranking model within LAIKA’s RAG architecture, optimizing inference with TensorRT-LLM and vLLM while simplifying runtime operations. Hosting Llama on company-controlled infrastructure across AWS and Scaleway ensured data security and compliance.
THEIR SUCCESS
Empowering vets to deliver better care, faster
Vets rely on LAIKA to accelerate clinical decision-making through rapid access to specialized expertise. The system analyzes cases up to 75x faster than human support, saving vets 5-10 minutes per case.Through LAIKA’s development, AITEM Solutions proved that AI delivers real impact when built around real-world use, backed by the right tools and developed in close collaboration with the people who use it. The team plans to expand this model across new healthcare applications.
75x faster than human support at analyzing critical cases
“LAIKA’s overall goal is to provide fast, accurate and specialized support for veterinarians in their clinical decision-making. Llama is a cornerstone to achieving that goal.”
Davide Tricarico - CTO, AITEM Solutions
“Having access to a large and active open-source community had a real impact. It allowed us to identify and overcome issues faster using best practices shared by others who were working on similar problems.”
Davide Tricarico - CTO, AITEM Solutions
“The Llama model family has played an important role in enabling these results. Not just because of its performance, but because of its open nature, flexibility and community support — all of which make it possible to build domain-specific, real-world AI tools.”
Shopify | Llama case studiesShopify uses Llama to generate product pages, localize content, and automate support, helping developers scale workflows and save time.Read more
ConsumerTech
Scribd, INC | Llama case studiesDelivering faster, cheaper and more accurate results with a Llama-powered AI content discovery assistant.Read more
Tech
Exati | Llama case studiesTransforming support for smart city platform clientsRead more
LAIKA offers easy access to specialized expertise when vets need it most.
Supported by Llama’s open-source community, AITEM Solutions rapidly iterated, applying best practices and feedback from over 100 veterinarians. Continuous fine-tuning and RAG optimization raised LAIKA’s response quality from 23% to 70% of answers rated four stars or higher.
This implementation leverages multiple instances of Llama and incorporates a reranking model developed by NVIDIA, which is available as an NIM container in NeMo Retriever.
5-10 minutes saved for every analysis
17.5x cheaper than human support
4-out-of-5-star ratings on average by clinical veterinarians
*All results are self-reported and not identifiably repeatable. Generally expected individual results will differ.
Stay up-to-date
Our latest updates delivered to your inbox
Subscribe to our newsletter to keep up with the latest AI updates, releases and more.
