Agile Data N’ Info Building Data Services with AI with Jason Taylor - Episode #79
Jan 9, 2026
In this engaging conversation, Jason Taylor, a former quant researcher now building AI-driven data services, shares insights from his journey into the startup world. He discusses the importance of broad skill sets in data careers and advocates for identifying roles based on strengths. Jason delves into how generative AI is reshaping data services and emphasizes caution in automated outputs. He also highlights the significance of domain expertise, the evolution of data collection, and the potential for AI to empower creativity in content creation.
AI Snips
Chapters
Books
Transcript
Episode notes
New Roles In Data Monetization
- There are three commercial data roles: raw collectors, integrators, and enrichers, and AI changes how value is created across them.
- Generative AI makes outputs non-deterministic, shifting product design toward process and quality controls.
Always Review AI Outputs
- Do implement a maker-checker review for AI outputs and always read or validate generated content before publishing.
- Treat AI work like human work: review, edit, and add deterministic checks to avoid costly hallucinations.
Disco: Automating Report Inventory
- Shane described building Disco to analyze 8,000 Cognos reports by extracting XML and using prompts to generate metadata and product canvases.
- They then used image-based LLMs to classify reports versus dashboards, saving huge manual effort.
