AI coding assistants help data teams build pipelines faster, but the last steps, like testing, fixing, deploying, and running smoothly, still slow things down for both people and AI.
Tower addresses this by offering a unified platform that combines storage, computing, and collaboration tools. It uses the Apache Iceberg format to integrate well with Snowflake, Databricks, and others, providing AI with fresh, company-specific data to reduce errors.
Tower, founded by former Snowflake engineers, has raised $6.4 million in pre-seed and seed funding. DIG Ventures led the pre-seed round, while Speedinvest led the seed round. Other investors include Flyer One Ventures, Roosh Ventures, Celero Ventures, Angel Invest, and angels such as Jordan Tigani of Motherduck and Olivier Pomel of Datadog.
The $6.4 million will help Tower grow its go-to-market efforts and enhance the platform. Since its launch, Tower has run over 200,000 jobs across more than 30,000 apps and has had 70,000 downloads of its Python SDK.
Foundational infrastructure for Vertical AI and SaaS builders
Serhii Sokolenko and Brad Heller, both former Snowflake engineers, started Tower because they saw a need for platforms that combine data processing with AI. They want to shift the focus from just writing code to delivering reliable production using real data.
Sokolenko shares with TFN, “After meeting at Snowflake, Brad and I started Tower in late 2024 because we saw a major shift happening in data engineering. For years, the field had been shaped by overly complex big data platforms. But suddenly, a new generation of data engineers was emerging – they were building with open-source tools and writing data applications directly in Python. The tooling had changed, but the infrastructure around them (Snowflake, Databricks, Fabrics, etc.) had not. “
“And this is a market challenge that’s become even more necessary to solve in recent months, thanks to the rise of AI-assisted software creation, such as Anthropic’s Claude. This has made it dramatically easier for data engineers to build their own data applications, but it has also made the missing infrastructure piece even more obvious. Once a data engineer has “Clauded” their data app, where does it actually run? The “last mile” – the production runtime of AI-assisted data engineering – still does not exist,” Sokolenko adds.
Tower combines analytics, storage, and processing with AI-human collaboration. It uses Iceberg tables to keep data ownership independent from vendors and provides fresh inputs that reduce AI errors. It also offers multi-tenant speed for fast iterations without relying on legacy cloud setups.
“Compared to the rest of the data ecosystem, Tower is easier to use, better integrated with the other tools developers have in their toolbox, and works natively with coding agents. Because Tower’s architecture is dramatically simpler and built on open technologies, it’s also substantially more cost-effective,” elaborates Sokolenko.
He adds, “Because it is so easy to get started, it’s especially exciting to see a new kind of user emerging that simply can’t use legacy solutions. People one would traditionally describe as “non-technical” (business founders, marketing managers, product managers) are now building pipelines, agents, and interactive dashboards – these are people who would not have touched this workflow a year ago.”
Unlike general cloud data warehouses such as Snowflake and Databricks, or AI coding tools such as GitHub Copilot and Cursor, Tower focuses on putting AI output into action. Competitors like dbt and Airbyte handle pipelines but don’t offer Tower’s AI-agent integration or focus on the final production steps.
What’s next?
With this funding, Tower plans to expand its go-to-market team, improve platform features for Vertical AI and SaaS, and become the go-to last-mile solution for data teams using Iceberg without extra operational hassle.
“Our vision is to turn Tower into a single environment where humans and AI agents can work together to turn AI-generated code and data pipelines into data systems that actually work,” concludes Sokolenko.
Currently, the team has over 12 nationalities, including Greece, Turkey, Sri Lanka, the US, Canada, and the UK.
