CEO & Co-Founder at Kepler, building deterministic AI infrastructure, starting with financial data.
Previously: Head of Business Engineering at Citadel, CTO at Veraset (acquired), Engineer at Palantir Technologies.
- Kepler: AI that never makes up numbers. We prevent LLMs from generating financial figures directly; instead, AI interprets intent while deterministic code retrieves data and performs calculations. After speaking with 137 financial firms, the pattern was clear: teams don't trust AI-generated numbers, and they shouldn't have to.
- Data infrastructure as the foundation for reliable AI
- Forward Deployed Engineering and closing the gap between software and real-world deployment
- Cost optimization in cloud and data systems, getting more signal per dollar
- Context engineering as a data pipeline problem, not a retrieval tuning problem
I write at vinoo.io about data engineering, AI infrastructure, and the craft of building production systems. Some recent posts:
I also publish on Substack (Efficiently) and contribute to the Kepler blog.
LinkedIn Learning Courses:
- Fundamentals of AI Engineering: Principles and Practical Applications
- Advance Your SQL Skills with dbt for Data Engineering
- Hands-On Introduction: Data Engineering
Conference Talks (selected):
- The Apache Spark File Format Ecosystem, Spark Summit / Data+AI Summit
- Accelerating Data Evaluation, Databricks Data+AI Summit
- Zero to Pipeline, O'Reilly Superstream Series
- Designing Data Pipelines, O'Reilly Live Training (3 sessions)
- Guaranteeing Pipeline SLAs and Data Quality Standards, Airflow Summit
- Rebuilt the official Apache Parquet website using Hugo
- Washington University in St. Louis, Computer Science, School of Engineering & Applied Science
- Member of McKelvey Engineering National Council and Alumni Board of Governors
- Advisor to Databand.ai, Horangi (cybersecurity), Sync Computing (Databricks optimization)