Tower raises €5.5 million to train data engineers in the age of AI

The Berlin startup, founded by two former Snowflake engineers, wants to be the platform where AI-generated data pipelines really work.

The hard part of building with AI is no longer getting the code. It is causing the code to run. That gap, between what an AI coding assistant can produce in minutes and what a production system actually needs to stay alive, is the problem Tower is trying to close.

tThe Berlin-based startup has raised 5.5 million euros (approximately 6.4 million dollars) through a seed and seed round, backed by investors such as Speedinvest and DIG Ventures, with angel participation from some of the most recognized names in the world of data infrastructure.

Tower was founded by Serhii Sokolenko and Brad Heller, both former Snowflake engineers who spent years watching engineers struggle not to write data pipelines, but to run them. Sokolenko, the CEO, previously worked in product management at Databricks and Snowflake in Berlin, and at Google Cloud, AWS and Microsoft in Seattle. Heller, the CTO, worked on Snowflake’s control plane. Tower is the third startup of both.

Its platform is designed to handle what the press release calls the “last mile” of AI-assisted development: the testing, debugging, delivery to production, and ongoing operation of AI-generated code. As Brad Heller says, the tool problem has changed.

“It’s easier than ever to write working code, but it’s still hard for humans, and even harder for AI agents to test it, troubleshoot it, deliver it to production, and operate it. That’s what we’re here to solve with Tower.”

Tower offers storage and compute on a single platform, built around the Apache Iceberg open table format. Iceberg has become the de facto open standard for analytical storage, supported by Snowflake, Databricks and most major data engine vendors, a deliberate choice that means Tower customers retain ownership of their own data and are not limited to a single stack.

The platform also supports AI agents that are fed new, company-specific data instead of the outdated public Internet files on which most base models are trained.

The pre-seed round was led by DIG Ventures and the seed by Speedinvest, along with existing investors. Additional backers include Flyer One Ventures, Roosh Ventures, Celero Ventures and Angel Invest.

The union of angels is notable: Jordan Tigani, CEO of MotherDuck and founding engineer of Google BigQuery; Olivier Pomel, CEO and co-founder of Datadog; Ben Liebald, vice president of engineering at Harvey; and Maik Taro Wehmeyer, co-founder and CEO of Taktile.

The list reads less like a random collection of investors and more like a who’s who of the generation of data infrastructure Tower is pitching to succeed, or at least scale. Tigani, in particular, has spent years arguing that the data industry was overengineered to achieve a scale it never really needed; Tower’s thesis that AI coding assistants have created a new problem of operational complexity fits neatly into that tradition.

Gaurav Saxena, director of engineering at Ford Motor Company, offered customer insight. Apache Iceberg, he said, represents genuine strategic value for businesses, but the operational demands of running it are a real limitation.

“Operating it effectively requires skills and ongoing maintenance that many data teams don’t have the staff for. The beauty of platforms like Tower is their ability to eliminate that operational overhead, making it much easier to adopt Iceberg without creating a dedicated internal team.”

The traction numbers in the press release, while early, suggest real usage. By February, a few months after launch, the platform had surpassed 200,000 runs on more than 30,000 unique applications and its Python SDK had reached 70,000 monthly downloads. These figures are self-reported and unaudited.

Sokolenko frames the company’s ambition in terms of where AI-generated production currently fails: not in generation, but in grounding.

“Builders can now create channels and agents in minutes, but they still need a platform that can reliably run them with real company data. Tower exists to turn those ideas into production systems, powered by information unique to each company instead of public and very outdated files from the Internet.”

Speedinvest’s Florian Obst, who led the company’s investment, pointed to multi-tenant architecture as a key differentiator: a platform designed from the ground up for rapid integration and iteration, rather than modernizing from an enterprise monolith.

Tower will use the new capital to grow its marketing team and deepen the platform’s capabilities. The market you are entering is competitive, Snowflake, Databricks and a wave of data infrastructure startups are investing heavily in the same AI-era data engineering story.

What Tower is betting is that none of them focus specifically on the problem that arises once the AI ​​finishes writing the code. tThat bet may be timely. The faster AI coding tools become, the greater the gap between what is generated and what is ready for production. Tower wants to be what fulfills her.