Functional AI product
Intake Engine
At a glance
- The problem
- The requirements-intake process Mike wanted to change was frequently criticized: ambiguous requirements led to repeated clarification, negotiation and rework.
- Mike’s role
- Mike conceived the product, designed the conversational experience and, by his account, built a functional application through AI-assisted development as an independent proof of concept. After it was brought in-house at Charter, he made bounded contributions to hardening it, while Design Technology and engineering owned assessment, review and merge decisions.
- The approach
- An AI persona interviews a subject-matter expert, asks contextual follow-up questions, notices gaps and produces a structured document, keeping working records of decisions, risks and deferred topics. Mike reports that he also designed how the product generates its directive instructions from criteria for a good answer.
- The outcome
- A functional application existed. Its deployed reliability, adoption and effect on users remain unknown, and whether it measurably improved requirements intake was not established.
Explore the work behind Intake Engine
Research findings, design decisions and project artifacts.

