An AI Team-Up: Snorkel AI and Microsoft Offer More to Azure Customers

By Alex Passett, Editor  |  April 26, 2023

Let’s talk Snorkel.

(In this case, though, we aren’t slipping into swimwear or even full wetsuits and diving in.)

Rather, this concerns Snorkel AI.

Snorkel AI is a data-centric AI company, with technologies developed at the Stanford AI lab with proven results from successes with Apple (News - Alert), Georgetown University’s CSET, Intel, Memorial Sloan Kettering Cancer Center (MSKCC), IBM, Uber, and even Google. Snorkel AI defines its technology as “AI beyond manual labeling, supercharged by foundation models.” With Snorkel AI, its customers and partners have experienced fewer errors via Snorkel-based applications, significant F1 improvements, a faster-than-hand-labeling generation of labels, faster application development for production-grade models, and other performance gains.

The long story short? Accelerated (News - Alert) time-to-value with a transformative approach to data-centric AI and meaningful use cases.

And now, the company tacks on another huge partner to its list:

Microsoft (News - Alert).

Earlier this month, Snorkel AI announced that it’s officially integrating Microsoft Azure AI services to speed AI development for Azure customers. In doing so, Snorkel AI has also joined the Microsoft for Startups program.

Snorkel AI’s inclusion in Microsoft for Startups reflects its efforts in data-centric AI and real-world, industry-spanning impacts. This news not only signals how Snorkel AI will benefit from access to Microsoft’s tech, market ecosystem and business support; it also enables the acceleration of ML Ops workflows with seamless integrations to Microsoft and Azure services, network, and advanced tools and platforms.

And notably, via integration with Microsoft Azure AI services to help data science and ML teams even more, Snorkel AI’s flagship programmatic labeling platform – Snorkel Flow – benefits greatly from this partnership. By using Azure Form Recognizer directly from Snorkel Flow, teams can operationalize PDF data for AI training set creation and model training (including OCR and layout parsing). With this, customers of Microsoft Azure are empowered to scale automation to, quote, “the iceberg-under-the-surface level of valuable enterprise data.”

“We launched Snorkel AI to bring a truly data-centric approach to AI development everywhere, and to make it practical for enterprises to build ML solutions for their most unique, highest value problems,” said Alex Ratner, Snorkel AI co-founder and CEO. “Now, we’re helping Microsoft Azure customers build models even faster by making it easy to deploy Snorkel Flow on Azure, train and serve models with Azure Machine Learning, and integrate directly with Azure AI services. We are honored to be chosen for Microsoft for Startups and to benefit from Microsoft’s experience, technology, reputation, and global customer base. Together, we will help more organizations realize the full potential of AI.”


Edited by Greg Tavarez
Get stories like this delivered straight to your inbox. [Free eNews Subscription]