AI Prototype Builder

A rapid-iteration engineer who collapses the gap between 'what if we built...' and a working AI demo that stakeholders can touch, break, and believe in — shipping proof-of-concept systems in days, not quarters.

- TESTBED methodology: thesis articulation through decision handoff in structured prototype sprints

- Model-task fit analysis with build-vs-API-vs-fine-tune decision matrices before writing code

- End-to-end scaffold: backend inference, lightweight API, interactive frontend wired in 48 hours

- Retrieval architecture design with vector store selection and hybrid search configuration

- Agent topology mapping with tool-use patterns, orchestration layers, debuggable fallback chains

- Evaluation harness construction with automated accuracy checks and comparison UIs

- Conversational interface patterns: streaming UX, turn management, confidence indicators, source attribution

- Human-in-the-loop workflow design with approval gates and feedback-as-training-data capture

- Assumption mapping and kill/scale decision packages with structured findings for leadership

Drop into Claude, ChatGPT, Cursor, or any AI tool. Bring your real prototype problem — a risky interaction pattern you need to validate, an architecture choice between retrieval systems, a user testing scenario you need to script. It thinks like an engineer who's shipped 50+ proof-of-concepts and knows which corners are load-bearing.

Bundles or prompts related to AI product development and rapid prototyping.