Tower Secures €5.5M to Empower AI-Era Data Engineers
▼ Summary
– Tower is a Berlin startup founded by ex-Snowflake engineers to solve the operational gap between AI-generated code and production-ready data pipelines.
– It has raised €5.5 million from investors like Speedinvest and DIG Ventures, with notable angel investors from the data infrastructure industry.
– The platform focuses on the ‘last mile’ of AI-assisted development: testing, debugging, deploying, and operating AI-generated code on real company data.
– It is built around Apache Iceberg for open data storage, allowing customers to retain data ownership and avoid vendor lock-in.
– The company’s early traction includes over 200,000 platform runs and 70,000 monthly Python SDK downloads, aiming to grow its team and platform capabilities.
A Berlin-based startup has secured €5.5 million in funding to tackle one of the most pressing challenges in modern data engineering: turning AI-generated code into reliable, production-ready systems. Founded by former Snowflake engineers, Tower is building a platform specifically designed to manage the testing, debugging, deployment, and ongoing operation of data pipelines created by AI assistants. The investment underscores a growing recognition that while AI can produce functional code rapidly, the operational complexity of running it at scale remains a significant bottleneck for organizations.
The company was founded by Serhii Sokolenko and Brad Heller, both veterans of Snowflake who observed firsthand the struggles teams face in operationalizing data workflows. Their platform centers on the Apache Iceberg open table format, a strategic choice that ensures customer data remains portable and avoids vendor lock-in with major engines like Snowflake and Databricks. By integrating storage and compute, Tower aims to provide a unified environment where AI-generated pipelines can be reliably executed.
Brad Heller, Tower’s CTO, explains the shift in focus. The primary challenge is no longer writing code, but ensuring it works in a live environment. AI agents and human developers alike struggle with the steps that come after the initial generation, testing for errors, fixing problems, and maintaining systems over time. Tower’s architecture is built to automate and streamline these post-generation tasks, effectively bridging the gap between a promising AI output and a stable production asset.
The funding round includes a pre-seed led by DIG Ventures and a seed round led by Speedinvest, with participation from Flyer One Ventures, Roosh Ventures, and several angel investors. The angel list features prominent figures from the data infrastructure sector, including Jordan Tigani of MotherDuck, Olivier Pomel of Datadog, and others, signaling strong industry validation for Tower’s approach.
Early adoption metrics, though self-reported, indicate promising traction. Within months of launch, the platform had processed over 200,000 runs across tens of thousands of applications, and its Python SDK saw substantial monthly downloads. This suggests a real demand for solutions that reduce the operational burden on data teams.
From a customer perspective, platforms like Tower address a critical pain point. Adopting powerful technologies like Apache Iceberg often requires specialized skills and continuous maintenance that many teams lack. By removing this overhead, Tower enables companies to leverage modern data formats without needing to build extensive in-house expertise.
The startup plans to use the new capital to expand its commercial team and enhance its platform’s features. It enters a competitive landscape where giants like Snowflake and Databricks are also investing heavily in AI-powered data tools. Tower’s unique bet is that the industry needs a dedicated solution for the “last mile” of AI-assisted development, the complex work that begins after the code is written. As AI coding tools become faster and more capable, the operational gap they create may only widen, potentially creating a substantial opportunity for the Berlin company.
(Source: The Next Web)
