Upriver
Category: AI Infrastructure
AI-native data engineering platform that automates enterprise data infrastructure for AI Upriver was founded in 2024. The company is led by Ido Bronstein. Based in San Francisco, California, United States. Team size: 11-50. Total funding raised: $18M. Latest round: Seed. Key investors include Valley Capital Partners, Hetz Ventures.
- Founded
- 2024
- Headquarters
- San Francisco, California, United States
- Team size
- 11-50
- Total funding
- $18M
Value proposition
Upriver's AI agent connects to an organization's full data environment (warehouse, orchestrator, and code) and autonomously manages the entire data engineering lifecycle — exploring data, building and validating pipelines, delivering reports, maintaining pipeline health, and encoding tribal knowledge — so data teams can deliver faster with reliable results.
Products and solutions
AI Data Engineering Platform — an autonomous agent system that handles: data exploration (ask anything about data environment), pipeline building (turn requests into validated pipelines), reporting (queries, analysis, reports from real data), maintenance (detects and fixes issues proactively), and knowledge encoding (tribal knowledge capture for new hire ramp-up). Integrates with Claude and Cursor AI development tools.
Unique value
The only AI agent purpose-built for the entire data engineering lifecycle with a continuously updated Data Context Layer that maintains a living map of the customer's full data environment — schemas, pipelines, lineage, query patterns, and semantic meaning of business logic — enabling autonomous execution of complex data engineering tasks without manual harness-building.
Target customer
Enterprise data engineering teams and data platform teams running modern cloud data stacks (Snowflake, Databricks, BigQuery, Airflow, dbt)
Industries served
Enterprise technology, e-commerce, advertising technology, financial services, media/publishing
Technology advantage
Agentic architecture with a Data Context Layer that maintains an always-updating map of the customer's data landscape, turning scattered tribal knowledge into an explicit, machine-usable model. Purpose-built validation harness ensures outputs are trustworthy. Pre-built integrations with Snowflake, Databricks, BigQuery, Airflow, dbt. Works within existing development workflows (Claude, Cursor).
How they differentiate
Unlike traditional data observability tools (Monte Carlo, Sifflet, Great Expectations) that only monitor and alert, Upriver's AI agent autonomously executes data engineering work end-to-end — building pipelines, fixing issues, and creating datasets without requiring manual intervention. Unlike AI coding assistants that generate SQL on demand (acting as autocomplete), Upriver maintains persistent context of the full data environment and proactively manages the data lifecycle.
Main competitors
Tamr, Saama, Bigeye
Key partnerships
Databricks, Snowflake, Nimble (web data enrichment), Technology partners include major cloud data platforms
Notable customers
Unity, DMGT (Daily Mail and General Trust), Nimble, Bright Insights, Bigabid, Resident
Major milestones
June 2026: Raised $14M seed round led by Valley Capital Partners and Hetz Ventures, Platform integrates with Claude and Cursor AI development tools, Partnerships established with Databricks and Snowflake
Market positioning
Positioned at the intersection of enterprise AI and data infrastructure, targeting the critical bottleneck of data engineering capacity that prevents AI projects from moving from proof-of-concept to production. Addresses the Gartner-identified problem that poor data quality and limited data availability are leading causes of AI project failures.
Geographic focus
United States (San Francisco/New York) with Israeli roots (founders are Israeli, backed by Israeli VC Hetz Ventures)
About Ido Bronstein
Led data infrastructure for Israel Military Intelligence (intelligence operations at scale); Tel Aviv University graduate
Official website: https://upriverdata.com