Tower secures $6.4M to solve AI deployment’s ‘last mile’
Tower, a startup founded by ex-Snowflake engineers Brad Heller and Serhii Sokolenko, has raised a $6.4 million funding round led by Speedinvest and DIG. The company is focused on fixing the “last mile” of Claude-powered data pipelines — the complex step of testing, validating and deploying AI applications into production environments.
Since its launch, Tower reports more than 200,000 runs on its platform, underscoring growing demand from teams struggling to move from promising prototypes to reliable, production-grade AI workflows.
Targeting the bottleneck in AI data pipelines
While many organizations have invested heavily in large language models and data infrastructure, the operational gap between experimentation and real-world deployment remains significant. Tower aims to close this gap by providing a unified environment where engineers and data teams can design, test, and roll out Claude-based and other LLM-driven applications.
The platform focuses on automating quality checks, managing configuration, and monitoring performance across complex data pipelines. This helps companies reduce failure rates, improve reliability, and accelerate time-to-market for new AI products.
Ex-Snowflake founders build for scale
Drawing on their experience at Snowflake, Brad Heller and Serhii Sokolenko are positioning Tower as a critical layer in the modern AI infrastructure stack. Their background in large-scale cloud data platforms informs the startup’s emphasis on observability, repeatability, and enterprise-grade governance.
Backers Speedinvest and DIG are betting that as more companies adopt Claude and similar models, the need for robust, testable deployment pipelines will only grow. With fresh capital and early traction, Tower is setting out to become a default choice for teams looking to industrialize their AI deployments rather than rely on ad hoc scripts and fragile workflows.