PhoenixAI has secured $80 million in Series B funding led by Sky9 Capital to accelerate development of its real time database infrastructure for autonomous AI agents. The capital will enhance product performance, expand enterprise operations, and strengthen data governance.
PhoenixAI (formerly CelerData), based in Menlo Park, California, announced an $80 million Series B funding round. The round was led by Sky9 Capital, with participation from Atypical Ventures, Olive Technology Ventures, and previous investors. This brings the company’s total known funding to around $100 million.
What is PhoenixAI?
PhoenixAI develops an AI native analytical database optimized for agentic AI workloads. It unifies real time streaming data (e.g., from Kafka) with historical lakehouse data (Apache Iceberg, Delta Lake) to deliver sub-second SQL queries at high concurrency, with “ingestion to queryable” latency under 5 seconds.
Key differentiators include:
- Vectorized execution and columnar storage for sub-second analytics on billions of rows.
- High concurrency (10K+ sustained queries).
- Enterprise features like SOC 2 compliance, fine grained access controls, row level security, and audit logging.
- Compatibility with existing stacks (Kafka, Flink, Spark, dbt, BI tools) without rip and replace, enabling fast production deployments (often 2-4 weeks).
The company rebranded from CelerData in May 2026 to emphasize its focus on the “Agentic AI Database.” It targets the shift from batch ETL pipelines (which introduce hours of latency) to real time systems needed for autonomous agents that issue unpredictable, high volume queries across live and historical data.
Customers in production include Coinbase (handling hundreds of billions of rows for fraud, compliance, and blockchain analytics), Demandbase, Conductor, Eightfold, AppLovin, and Fanatics. These users report benefits like sub-second latencies on complex joins, cost reductions (e.g., 90% in one case), and support for agentic workloads on petabyte scale data.
How will PhoenixAI use the funds?
The funding will accelerate:
- Product development of the AI native database.
- Go to market expansion.
- Deepening governance and security features for regulated industries (finance, healthcare, etc.).
The raise aligns with the rapid move of agentic AI from prototypes to production. Agents require low latency access to fresh, unified data at massive scale, capabilities that traditional warehouses (batch oriented) or even some real time OLAP tools struggle to provide without heavy pre modeling. PhoenixAI positions itself against Databricks (real time vs. batch first), ClickHouse, and Snowflake (freshness vs. latency).
Recommended: Supabase Raises $500 Million In Series F Funding Round
Broader tailwinds include explosive growth in AI infrastructure needs, with agentic systems driving demand for databases that handle unpredictable query patterns, high concurrency, and seamless real time/historical integration. Enterprise adoption in sectors like finance, retail, logistics, and manufacturing further supports this.
Leadership and Investors
Rick Underwood, President, brings relevant experience as former CEO of Clumio (acquired by Commvault) and senior GTM leadership at Snowflake during its IPO. He oversees go to market and operations.
Sky9 Capital (lead) focuses on AI, fintech, and deep tech. It has backed companies like Moonshot AI and WeRide. The firm views the database layer as central to the agentic AI infrastructure shift.
This Series B validates PhoenixAI’s pivot and technology in a competitive real time analytics and AI data infrastructure market. The rebrand, executive hire, and substantial raise signal confidence in its ability to capture value in the foundational AI stack. Strong early customer traction with high profile enterprises demonstrates product market fit for production agentic use cases. Challenges ahead may include scaling sales, maintaining performance at even larger scales, and competing with incumbents expanding into real time capabilities. Overall, the funding equips PhoenixAI to push boundaries in real time, governed data infrastructure tailored for the next wave of AI agents.
Please email us your feedback and news tips at hello(at)techcompanynews.com
