Sphinx, a company dedicated to advancing AI for data, has launched with a $9.5 million Seed round and the debut of its AI copilot for data professionals. In addition to officials from Databricks, Windsurf, and Together AI, Lightspeed spearheaded the investment round with backing from prominent business titans like Steve Cohen and Naveen Rao. Bessemer Venture Partners, Box Group, K5, Impatient VC, and others also participated. The funding will support the creation of agentic AI that is native to data and data science processes.

While AI has become an essential part of software development, data science has not yet reaped the same benefits. Data science workflows are much more exploratory and iterative than coding approaches. Sphinx’s Copilot is designed specifically for these challenges, emphasizing accurate data representation, model verification, and evidence-based insights rather than simply generating code or conclusions.

“AI’s role in data analysis is entering a new era, and Sphinx is at the forefront,” said Bucky Moore, partner at Lightspeed. “By focusing on the workflows of data teams, Sphinx’s agents will increasingly take on the labor-intensive tasks of deriving insights. With enterprises under pressure to harness their data for smarter decisions, Sphinx delivers speed and precision that are truly transformative.”

The Sphinx Copilot, available now, collaborates with users to turn raw information into actionable insights through autocomplete and agent-driven reasoning. It can refine forecasts, optimize operations, and drive applications from supply chain planning to sports analytics. Built to integrate seamlessly with environments such as Jupyter notebooks and VSCode, it meets data professionals where they already work.

The CEO and co-founder of Sphinx, Rohan Kodialam, claims that “data is still underserved, despite AI having revolutionized natural language and code”. “Our mission is to innovate at the interface of AI and data, creating real impact for industries like consumer goods, retail, and financial services.”

Sphinx strikes a balance between exploration and value extraction by emphasizing reasoning models and the interpretation of structured and semi-structured data. Its copilot is positioned to serve over 93 million Jupyter users and address a $100 billion market for data insights. Early adopters are already seeing results—Brian Tate, CEO of Oats Overnight, noted that the tool allowed his data scientists to uncover shopper behavior patterns in minutes instead of hours.

Founded by Kodialam, formerly an AI research leader at Citadel, and Jamie Bloxham, an early technology leader at MosaicML, Sphinx emerged from their shared experience of seeing powerful AI models disconnected from real-world data workflows. Their goal is to bridge that gap with research and products that accelerate and elevate the work of data practitioners.