The LLM Mesh acts as a secure API gateway to break down hard-coded dependencies, managing and routing requests between applications and underlying services.
Dataiku partners with leading providers of large language models (LLMs), vector databases, and accelerated and containerized compute capabilities. This includes Hugging Face for serving production-grade local LLMs easily and efficiently.
SEE AVAILABLE LLM CONNECTIONS
Should you build your GenAI applications or AI agents on top of today’s best model? Or should you hold out just a little longer until a new one emerges that’s more accurate, powerful, or better suited for your use case?
With Dataiku, you don’t have to choose. Balance trade offs between cost, performance, privacy, security, regulatory requirements, and more with a multi-LLM strategy. The LLM Mesh is the most comprehensive and agnostic LLM gateway offering on the market, partnering with the top Generative AI players for secure access to thousands of LLMs (including third-party or self-hosted).
We compared five different LLM models, something we couldn’t do without Dataiku, and found that one model continued to outperform the rest by a healthy margin for our specific use case.
We realize that no single LLM is good for every use case.
Deepthi Sanam
Group Lead of Data Engineering at Novartis
Kenan Yates
Group Product Manager at System1Analyze up-to-the-minute data on LLM usage via Dataiku Cost Guard to ensure GenAI and AI agent applications aren’t breaking the bank.
A fully auditable log of who is using which model or service and for what purpose allows for cost tracking and internal re-billing.
DIG DEEPER ON LLM GUARD SERVICES
When it comes to screening GenAI or agentic applications for private data or inappropriate content, the LLM Mesh — via Dataiku Safe Guard — can evaluate requests and responses for sensitive information. This includes detecting personally identifiable information (PII), toxic content, or forbidden terms.
The LLM Mesh then takes appropriate action: redact this sensitive information before sending the request to the LLM API, block the request entirely, and/or alert an admin.
LEARN MORE ABOUT LLM GUARD SERVICESDataiku Quality Guard provides standardized, enterprise-grade quality assurance for LLMs and Generative AI applications.
Automatically compute standard evaluation metrics like LLM-as-a-judge techniques (answer relevancy, answer correctness, context precision, etc.) as well as statistical techniques (BERT, Rouge and Bleu, etc.).
DISCOVER LLM GUARD SERVICES
Andres Buser
Chapter Lead New Modality IP at Roche
Qualifying, documenting, and framing your organization’s use of LLMs is a critical step toward regulatory readiness.
Keep model documentation up to date as well as rationalize which models should (or should not) be used for what use cases with the Dataiku LLM Registry.
MORE ABOUT AI GOVERNANCE WITH DATAIKUThe LLM Mesh monitors the round-trip performance for LLM services and providers so teams can diagnose issues and select the optimal service based on application needs and SLAs.
Additionally, caching of responses to common queries avoids the need to regenerate responses, offering both cost savings and a performance boost.
Dataiku brings not only the LLM Mesh, but all the tools required for builders across the business to bring GenAI and AI agents to life — fast. .
The Universal AI Platform™ provides your stakeholders with everything they need to create GenAI applications and AI agents at scale.
DISCOVER GENAI & AGENT BUILDER TOOLS
Johnson & Johnson’s Vision team partnered with Dataiku to conduct a two-day GenAI training and hackathon event. The hackathon was a profound success, producing working prototypes spanning vision care and surgical vision, including use cases for consumers, sales, contracts, and customer insights.
READ J&J'S STORY
Novo Nordisk leverages generative AI features and capabilities in Dataiku to create predictive medical insights, automate scientific literature summaries, improve incentive compensation understanding through real-time insights for field sales teams, and much more.
READ NOVO NORDISK'S STORY
With over 15 million reviews total and 10,000+ new reviews coming online each week, Whataburger was dealing with a lot of unstructured data. With Dataiku, they quickly built a generative AI solution that is entirely visual, code free, and easy to manage.
READ WHATABURGER'S STORY
For Novartis, the LLM Dataiku Mesh abstracted intricate model configurations — like model parameters and prompt formats — into holistic, user-friendly UI options, thereby streamlining the experimentation process with various models across multiple service providers.
READ NOVARTIS' STORY
The Dataiku LLM Mesh allowed System1 to seamlessly integrated large language models (LLMs) into their workflows, providing business users with end-to-end visibility into the data transformation process — from raw data inputs to the prompts used and, ultimately, to the final AI-generated outputs.
READ SYSTEM1'S STORY
A team of patent attorneys at Roche developed a robust Dataiku application that executes a complex search and retrieval flow for comprehensive case law analysis, significantly improving efficiency — all with visual tooling and without the need for extensive IT support.
READ ROCHE'S STORY
