AI cold-call simulator
A voice-AI practice partner. Junior agents run live cold-call scenarios and get structured, scored feedback from a custom LLM scoring engine.
- ElevenLabs
- LLM scoring
Each one started as a problem a team brought to me.
A voice-AI practice partner. Junior agents run live cold-call scenarios and get structured, scored feedback from a custom LLM scoring engine.
About 40 reusable Claude skills and automations shared across departments, plus 80+ internal skills, project specs and instruction sets.
A custom AI training curriculum and parts of an internal learning management system, used to train 50+ employees.
Apps for marketing content creation and operations workflows, taken from prototype to daily use.
Pulls token costs from the AI platform’s admin console into a live web app. Each employee sees their own spend, and admins can filter by person.
An animated, interactive walkthrough of Value Stream Science’s AI agent platform for research scientists.

Scientists are the experts on their research, but turning that knowledge into computation usually means hiring a computational specialist. Methods interviews a researcher about how their work runs, maps it into standardized protocols and templates, and shows which steps can become automated analysis workflows, with the scientist in charge throughout.
I designed and built the interactive demo that walks audiences through the product from first conversation to finished workflow.
A closer look at how I run the AI portfolio, track what it costs, share what works and support the people using it.