This is a guest blog post by the Ryght team.
Who is Ryght?
Ryght is building a corporate-grade generation AI platform tailored for the healthcare and life sciences sectors. Today is the official release of the Ryght preview, and now everyone is open to the public.
Life sciences companies accumulate a wealth of data from diverse sources (lab data, EMR, genomics, claims, pharmacies, clinical, etc.), but the analysis of that data is archaic and requires a large team of all sizes, from simple queries to developing useful ML models. There is a great demand for practical knowledge to promote drug development, clinical trials and commercial activities, and the rise of precision medicine will only drive this demand.
Ryght’s goal is to ensure life science experts can quickly and safely gain the insights they need. To that end, they are building a SaaS platform that provides industry-specific AI Copilots and custom built solutions for professionals and organizations to accelerate research, analysis and documentation across a variety of complex data sources.
Recognizing the fast-paced and constantly changing AI landscape, Ryght embraced his face as a technical advisory partner early on in the journey through his expert support program.
Together we will overcome the challenges
Hugging Face’s partnership with expert support has played a key role in fostering the development of generator AI platforms. The rapidly evolving landscape of AI can revolutionize our industry, with the highly performance of faces and enterprise-ready text-generating inference (TGI) and text-embedded inference (TEI) services being a game changer in itself. – Johnny Crupi, CTO of Ryght
Ryght faced several challenges as he set out to build a generative AI platform.
1. The need to quickly expand your team and provide information in a highly dynamic environment
With AI and ML technology advancing so quickly, it is important that teams stay on top of the latest techniques, tools and best practices. This continuous learning curve is steep and requires collaborative efforts to provide information.
Once you have access to Face’s team of experts operating at the heart of the AI ecosystem, you can keep up with the latest developments and models related to your domain. This is achieved through open and asynchronous communication channels, regular advisory meetings and dedicated technical workshops.
2. Identify the most (cost)effective ML approach in the noisy sea of options
The AI field is bustling with innovation, with a wealth of tools, libraries, models and methodologies. For startups like Ryght, it is essential to get through this noise and identify which ML strategies are most applicable to the unique use cases of the life science sector. This includes not only understanding the cutting edge of the day, but also taking precedence on which technologies are relevant and scalable for the future.
The embrace face acts as a partner for Ryght’s technical team. Helps in designing solutions, proof-of-concept development, and optimizing production workloads. This includes customized recommendations for libraries, frameworks, and models, as well as demonstrable examples of the best models for your specific needs of Ryght, as well as how to use them. This guidance ultimately streamlines the decision-making process and reduces the time to development.
3. Requirements for developing performance solutions that emphasize security, privacy and flexibility
Given its focus on enterprise-level solutions, Ryght prioritizes security, privacy and governance. This requires a flexible architecture that can interface with a variety of large-scale language models (LLMs) in real time. This is an important feature for life science-specific content generation and query processing.
By understanding rapid innovation within the open source community, especially with regard to medical LLM, they embraced an architectural approach to supporting “pluggable” LLM. This design choice allows for seamless evaluation and integration of new or specialized medical LLMs as they emerge.
On Ryght’s platform, each LLM is registered and linked to one or more customer-specific inference endpoints. This setup not only protects your connections, but also provides the ability to switch between different LLMs, providing unparalleled flexibility. This is a design choice that is made possible by the adoption of Face’s Text Generation Inference (TGI) and inference endpoints.
In addition to TGI, Ryght has integrated text-embedded inference (TEI) into the ML platform. The provision of open source embedding models using TEI is significantly improved over relying solely on proprietary embeddings. This allows Life to have faster inference speeds, eliminate rate limiting concerns, and provide a unique fine-tuning model tailored to the unique requirements of the Lifescience domain.
When catering multiple customers simultaneously, their systems are designed to handle large numbers of simultaneous requests while maintaining low latency. Their embedding and inference services go beyond simple model invocations, covering a set of services well-versed in batches, queuing and distribution of model processing across the GPU. This infrastructure is important to avoid performance bottlenecks, prevent users from experiencing latency, and thereby maintain optimal system response times.
Conclusion
Ryght’s strategic partnership and integration with Face’s ML services underscores our commitment to providing cutting-edge solutions for healthcare and life sciences. By adopting a flexible, secure and scalable architecture, they ensure that the platform remains at the forefront of innovation, providing clients with unparalleled service and expertise to navigate the complexities of the modern medical domain.
Sign up for Ryght Preview. Now, Life Sciences knowledge workers are publicly available as a free, secure platform with frictionless onboarding. Ryght’s Copilot Library consists of a diverse collection of tools for accelerating information search, synthesis and structuring of complex, unstructured data, and get what could have taken weeks to complete days or hours. To inquire about custom builds and collaboration, contact our team of AI experts to discuss the Ryght company.
If you would like to know more about hugging face professional support, contact us – our team will reach out to discuss your requirements!
