The AI copilot and agent platform for the teams who actually run the business. Connect your tools and data, automate the workflow, and turn the work that repeats into agents that don't clock out.
Most companies guard their original idea like an heirloom. Continual did something rarer: it kept the name, kept the founders, kept the conviction that AI should disappear into the work - and quietly replaced the product underneath. In 2021 it sold predictions in your data warehouse. Today it sells agents that run your operations. Same marquee, different show.
The throughline is a stubborn idea: building with AI shouldn't require an archaeology dig through your own infrastructure. Continual's pitch is plain enough to fit on a sticky note - connect your tools and data, automate workflows across agents and teams, and convert the processes that repeat into agent-powered applications. The unglamorous middle of a business, finally automated by something that learns.
That line, from CEO Tristan Zajonc, reads differently now than it did in 2022. Back then it described a predictive-modeling platform on the modern data stack. Read it today and it sounds like a roadmap for the agent era the company walked straight into.
Wire Continual into the apps, APIs, and data your team already runs on. The agent sees what your people see.
Orchestrate work across agents and humans together - the busywork handled, the judgment calls escalated.
Turn a repeatable process into a production-ready, agent-powered application your users can actually rely on.
Earlier in its life, the same philosophy showed up as operational AI on the modern data stack: analytics teams used their existing SQL and dbt skills to build continually-improving models - customer churn, inventory forecasts, risk - directly on warehouses like Snowflake. No ML PhD required. The audience was always the people who run the business, not just the people who tune the models.
Spent a decade in the trenches of machine-learning infrastructure. His first startup, Sense, was an early enterprise ML platform acquired by Cloudera in 2016, where he went on to serve as CTO for Machine Learning. He sets Continual's direction - and clearly enjoys reading the market a beat early.
Built RichRelevance, one of the world's leading personalization providers, before it was acquired by Manthan in 2019. The kind of engineer who has already shipped recommendation systems at internet scale - now pointing that experience at agents.
The Series A was led by Innovation Endeavors, with a roster that doubles as a modern-data-stack hall of fame.
Continual comes out of stealth to bring AI to the modern data stack - declarative, automated, warehouse-native.
The platform exits beta. In the prior quarter it doubled active users, deployed models, and booked ARR.
Zajonc previews Continual as "the AI copilot platform for SaaS applications" - a developer platform for generative AI.
Public reintroduction as the AI copilot platform for applications, built for the agent era.
The platform settles into its current shape: connect, automate, ship - agents working alongside teams.
"Run your operations with agents."
"A world where intelligent agents work alongside every person and team to make the impossible possible."
Talks, demos, and primary sources. (Search links open the most relevant recent results.)