Research
We’re looking for an AI Researcher to join our team.
AI does not natively interface with data. This is the key problem we’re working on at Sphinx, and it broadly splits into two areas:
- Representation Learning: How can we convert data, from small arrays of floats to multi-terabyte tables in data lakes, into context that AI can ingest and understand
- Agentic Learning: How can we get agents to efficiently explore and exploit across various data modalities to find answers that maximize hard-to-define metrics (like the overall business value of a data insight)
We’re looking for researchers who are excited by these challenges, and who want to aggressively innovate and drive impact. Our research team works in close collaboration with our engineers – not in the abstract, but in the sense that innovations make it to prod within one day – and we consider our entire pipeline as part of the agentic environment we seek to optimize
Recent work from our research team include:
- Developing benchmark-beating agents for data science using kernels as long-term memory
- Building methods for effective data sampling using LLMs
- Innovating on a dynamically evolving semantic ontology for enterprise data (think Palantir, but with zero onboarding)
We’re looking for research experience, understanding of both AI/ML research and traditional data/statistics, and a keen drive to see research drive impact.
About Sphinx ML, Inc.
Sphinx AI (www.sphinx.ai/) builds frontier AI agents for data science. Our agents combine statistical algorithms, multimodal understanding, and business logic to accelerate the discovery of value from data. Sphinx's copilot integrates seamlessly into most data science IDEs (Jupyter/VSCode), compute layers, databases and data lakes. We augment every step of the data science process from data discovery, to exploratory analysis, to modeling, to impactful data-driven storytelling.