June 8, 2022
Today, we’re excited to announce the general availability of Continual, the missing AI layer for the modern data stack. We’ve also raised a $14.5M Series A, led by Innovation Endeavors and joined by Amplify Partners, Illuminate Ventures, Inspired Capital, Data Community Fund, Activation, New Normal, GTMfund, and angels Tomer Shiran, the founder of Dremio, and Tristan Handy, the founder of dbt Labs. This follows on our announcement of our public beta and partnerships with Snowflake and dbt Labs. While we have lots still in store, we encourage everyone to request a demo to experience Continual for yourself.
Continual is an operational AI platform for the modern data stack. If you’re new to operational AI or have ever been frustrated by the complexity of today’s ML tools and MLOps solutions, Continual is for you. Continual enables data teams to build continually-improving predictive models — from customer churn to inventory forecasts — directly on top of cloud data platforms like Snowflake, BigQuery, Redshift, and Databricks. Unlike traditional MLOps platforms, Continual is powered by a declarative design and end-to-end automation. There’s no complex engineering or operational burden.
Since launching our public preview in December, we’ve been thrilled to see the breadth of use cases customers are bringing to Continual. Companies like Enverus and Aurora Solar are using Continual to better predict customer behavior and identify growth and cross sell opportunities. In the retail industry, customers like Veronica Beard are using Continual to maintain models to help optimize marketing efforts and forecast store sales given changing market dynamics.
We believe that the easier it is to build and maintain predictive models, the more impact they will have across every business. Continual is our effort to simplify operational AI and thereby unlock its true potential.
We founded Continual after a decade of working on traditional MLOps platforms and being frustrated with the results. While today’s leading enterprises are increasingly AI-driven, operational AI is drowning in complexity. Production machine learning platforms bear a striking resemblance to the early days of “Big Data”, where writing map-reduce jobs was required to answer simple analytical queries. Even for the simplest use cases, putting models into production requires complex infrastructure, bespoke pipeline jungles, and never ending operational babysitting. It’s costly, slow, and just no fun.
Over the last decade, data platforms have undergone a profound transformation driven by the rise of cloud data platforms like Snowflake, BigQuery, Redshift and Databricks. These platforms have struck at the heart of Big Data complexity and rightfully become the center of gravity for modern data architectures. Building upon this new potential, a vibrant modern data stack ecosystem has emerged spanning data ingestion, transformation, and activation. These modern data stack tools deliver best-of-breed solutions to common parts of the data lifecycle.
Continual brings this same ethos of radical simplicity to operational AI. It sits directly on top of cloud data platforms and enables anyone to build and maintain predictive models that never stop learning from their data. With Continual, putting a predictive use case into production, including model maintenance and monitoring, requires no custom pipelines or infrastructure. Whether you’re an analytics engineer or data scientist, you can focus on what matters most — delivering business impact.
Continual’s secret sauce is its declarative approach to operational AI. Unlike traditional MLOps platforms which require bespoke pipelines for each business use case, Continual cleanly separates the declarative modeling layer from the imperative operational layer. This approach is inspired by next-generational declarative ML platforms at Apple, Facebook, Google, Uber, and Stripe.
This separation of the declarative and operational layers allows anybody to build production ML use cases without worrying about the intricacies of the operational pipeline or infrastructure. It also allows Continual to support multiple interfaces, from a UI for quick exploration to first-class dbt integration for SQL-centric production workflows, all backed by the same underlying abstractions.
The declarative layer radically simplifies building new ML use cases and the operational layer is what makes the whole process tick. Continual has built in automated capabilities to solve the most common ML problem types we see from customers. But we've just scratched the surface of what's possible with this split design. In the coming months, we plan to announce additional built in ML capabilities and extensibility of the operational layer to allow data scientists to bring custom models and capabilities to Continual. If you’d like to join our early access program for these features, please reach out.
Our vision for Continual is to unite analytics and AI teams on a shared platform that simplifies operational AI for everyone. For AI to fulfill it’s true potential, the complexity of today’s MLOps solutions must give way to simpler and higher-level approaches that empower all users and place the emphasis on what really matters — delivering business value at scale across a business.
If you’re interested in learning more about Continual, you can:
Let’s say goodbye to complexity and bring AI to the modern data stack!
Traditional AI platforms require users to manage complex data infrastructure and write bespoke ML pipelines for each case. We let users build predictive models using a declarative workflow that radically streamlines operational AI at scale.