Introducing Application Studio and Announcing Our $35M Series B Funding
Over the past year, we’ve worked hard to deliver Snorkel Flow, the first AI platform to provide all the power of machine learning without the pains of hand-labeling. Snorkel Flow lets you label data programmatically, train models flexibly, improve performance iteratively, and deploy AI applications quickly. We are incredibly proud of the value that our customers, including two of the three largest banks in the US, several government agencies, and global enterprises in insurance, telecommunications, biotech, and more, are deriving by using Snorkel Flow, including 10x-100x faster development times and millions of dollars of ROI.
Today, we are thrilled to announce a major step towards making AI development even faster with Application Studio, a visual builder with templated solutions for common AI use cases and easy construction of new and custom ones. Application Studio is currently in preview and will be generally available later this year within Snorkel Flow.
Since our launch out of stealth last July, we have received an incredible amount of interest in Snorkel Flow from all market segments. Hundreds of data scientists, developers, product managers, and business leaders from the Fortune 500, global enterprises, public sector, IT consulting firms, AI startups, and academia have joined the waitlist for Snorkel Flow. To meet this overwhelming market demand and make AI practical for all, we’re excited to announce that we’ve raised $35M in Series B funding, led by Lightspeed Venture Partners. We’re looking forward to using this investment to continue expanding Snorkel Flow’s capabilities, bringing innovation to the market, and growing our team.
Sign up for a Snorkel Flow demo here, and if you are on our waitlist already, we will be getting in touch with you soon!
For academic researchers and data scientists at qualified non-profit organizations interested in early access to Snorkel Flow: as of today, you can apply for a no-cost license via our Research Accelerator pilot program! See below for more details.
Please join us on April 13th at 9 am PDT to learn more about Application Studio, our Series B funding, and the Research Accelerator pilot.
AI Beyond Hand-Labeling
Snorkel AI’s journey began at the Stanford AI lab in 2015, where the Snorkel AI founding team started studying the then largely overlooked problem of labeling and managing the training data that machine learning models learn from. Most AI approaches work by learning from thousands or millions of examples of a task that have been labeled with the correct answer or action–known as “training data.” The not-so-hidden secret of AI is that even today, this training data requires vast volumes of painstaking manual labeling effort.
We began to see then that this ‘dark ages’ approach to labeling training data by hand would not scale, especially as machine learning approaches became increasingly data hungry, and in verticals like healthcare, government, finance, etc., where data is incredibly difficult and costly to label due to privacy, expertise, and frequent re-labeling requirements. Indeed, training data has arguably become the critical bottleneck in AI today.
Convinced that there had to be a better way than hand-labeling, we spent over four years developing new programmatic approaches to labeling, augmenting, structuring, and managing this training data, co-developed and deployed with some of the world’s leading organizations like Google, Intel, Apple, and Stanford Medicine, and represented in over forty peer-reviewed publications. With this new programmatic approach, you could create massive amounts of labeled training data programmatically in a matter of hours, instead of weeks or months labeling data by hand–unlocking a fundamentally new, faster, and more practical way to develop AI.
To make this new approach a usable reality, we spun out of the Stanford AI lab in 2019 to build Snorkel Flow, the first AI application development platform powered by programmatic data labeling. Snorkel Flow makes it possible for data scientists, developers, and subject matter experts to collaboratively create and manage training data rapidly, train custom ML models, analyze and iterate to drive systematic improvements, and adapt and deploy AI applications quickly.
Introducing Application Studio (in Preview)
Adopting a programmatic approach to training data, of course, does not alone turn AI application development into a turn-key process. For different data modalities, ML task types, and use case settings, there is a range of critical components and best practices–including various strategies, data visualizations, model classes, application graphs, error analyses, and more–that make the difference between success and failure.
We are thrilled to announce the launch of Application Studio, a new capability within Snorkel Flow that templatizes, productizes, and organizes the best practices and components we’ve developed over years of engagements with hundreds of organizations and partners. The result is the fastest way to build AI applications without hand-labeled training data–and even without code using our no-code UI–while still maintaining full customizability for expert data scientists and engineers.
With the introduction of Application Studio, you can develop AI applications faster than ever before. Application Studio provides data scientists, developers, and subject matter experts with:
- Pre-built Application Templates – Based on industry-specific use cases such as contract intelligence, news analytics and customer interaction routing or common AI task types such as text and document classification, named entity recognition, and information extraction, templates give enterprise data science teams a head start in developing their applications. Packaged application-specific pre-processors, programmatic labeling templates, models, and features make customizing applications as easy as dragging and dropping new logic into the application flow.
- High-Performance Models – Enterprises can use their private data, labeled programmatically, to train state-of-the-art, open-source models available in the platform or outside of it via Snorkel Flow’s Python SDK. Programmatic labeling replaces weeks or months of costly hand-labeling yielding highly accurate model performance.
- Collaborative Workflows – Application Studio allows for an intuitive decomposition of complex applications into modular parts so that data scientists, developers, and subject matter experts can collaborate easily and efficiently.
- Auditable and Adaptable Capabilities – With Application Studio, the entire pipeline from training datasets to user contributions is easily versioned, audited, and adapted to new data or objectives. Unlike other application platforms that rely on hand-labeled data, there is no need to start from scratch.
- Data Privacy at Enterprise Scale – Data privacy constraints are often among the most significant blockers to AI development. With Application Studio, training data labeling and management can be kept in-house and done without humans needing to view most of the data—setting a new high bar for practical, private machine learning.
Announcing our $35M Series B Funding
We’re excited to announce that Snorkel AI has raised a $35M Series B funding round led by Ravi Mhatre and Raviraj Jain at Lightspeed Venture Partners. This round also saw participation from previous investors Greylock, GV, In-Q-Tel, and Nepenthe Capital and new investors Walden and funds and accounts managed by BlackRock. It follows our previous $15M of Seed and Series A round financing we raised last year.
Our investors are world-class experts and ex-founders of companies focused on enterprise software, developer tools, and AI/ML with previous investments in Appdynamics, Mulesoft, Rubrik, and many more. We are immensely grateful to them for the help they’ve provided every step of the way. This round of financing will help us grow the team to pursue our vision of making AI practical for every organization.
Apply for Research Accelerator Pilot
In keeping with our roots and years of efforts from the Stanford AI lab, Snorkel AI encourages academic and non-commercial innovation with support for faculty, researchers, students, and data scientists at qualified non-profit organizations. The Research Accelerator is a pilot program for researchers to receive a grant for free and early access to Snorkel Flow. If you are interested, apply here.
We are just getting started
We have big plans for Snorkel AI. We’re building what we believe will be the practical and scalable default for AI teams to build intelligent applications. We’re investing in new features that will give you out-of-the-box solutions along the entire ML lifecycle, including:
- Workflows specially designed for Subject Matter Experts
- Richer model evaluation tools
- Greater governance and auditability
- Continuous monitoring capabilities
- And a few things we aren’t announcing just yet. 😉
Join us on 13th April 2021 at 9 am PDT to learn more about Application Studio, our Series B funding, and the Research Accelerator Pilot.
Join our Team
If you’re passionate about solving problems nearly every data science and developer team struggles with and want to shape the future of AI, we want to hear from you! We’re hiring for engineering, SRE, product, design, marketing, sales, solution engineering, and many other roles. Check out our job opportunities for more details.