Upriver secures $14 million to address the enterprise data challenges in AI.

      Most enterprise AI initiatives do not fail due to poor models. They fail primarily because the data that supports them is disorganized: malfunctioning pipelines, incompatible systems, and context confined to the knowledge of a single engineer.

      Upriver, an Israeli startup, has secured $14 million to automate the data cleaning process, believing that this tedious yet essential aspect is where the success or failure of the AI era will truly be decided. The seed funding was led by Valley Capital Partners and Hetz Ventures.

      Noteworthy is the list of angel investors, which includes prominent figures in the data-tooling realm: Lew Cirne, the founder of the observability leader New Relic; Abe Gong from the data-quality initiative Great Expectations; and the creators of the Israeli data-security startup Cyera.

      Upriver claims that its platform is already being utilized by Unity and the media organization DMGT, and it collaborates with Databricks and Snowflake.

      Positioned as an “AI data engineering platform,” Upriver acts as an agent that integrates with a company’s entire data infrastructure, including Snowflake, Databricks, BigQuery, Airflow, and dbt. It explores, builds, and validates data pipelines, rectifying issues when they arise, while also capturing the “tribal knowledge” typically held by individuals.

      The expectation is that data engineers will no longer spend their time troubleshooting faulty pipelines and manually connecting disparate tools, but rather focus on interpreting the data's meaning.

      The founders bring a unique perspective: Ido Bronstein, the CEO, and Omri Lifshitz, the CTO, spent ten years developing large-scale intelligence systems—work reported by Business Insider to have been conducted for the Israeli military—before realizing that every modern cloud-based company faces similar challenges.

      The recent funding will be invested in engineering, sales efforts, and enterprise implementations.

      This timing aligns with a broader industry recalibration. Following two years of investment in models and hardware, companies are now examining the actual returns from AI, frequently finding that the technology falters due to poor data. Numerous startups, from Capsa in private equity to Upriver within the data infrastructure, are promoting a common message: reliable, clean data is critical for transitioning an AI initiative from concept to successful execution.

      While the $14 million seed round is early and unproven at scale, the belief that foundational elements are more crucial than the model itself is gaining traction among investors.