Engine

Builds and ships entire products alone using AI tools. CEO, CTO, designer, and marketer in one. Uses Claude Code, Cursor, and similar tools to ship at the speed of a team.

Builds, configures, and maintains AI agents using platforms like Salesforce Agentforce or Microsoft Copilot Studio. Defines agent behaviors, connects data sources, sets guardrails.

Defines what AI features should do and should not do. Writes eval criteria instead of acceptance criteria. Manages the tension between model capability and user expectations.

Sets AI strategy at the company level. Model selection, build vs buy decisions, AI governance, training programs, vendor negotiations, ROI measurement.

Systematically attacks AI systems to find failures before users do. Prompt injection, jailbreaking, bias testing, hallucination mapping, safety boundary testing.

Designs systems that give AI the right information at the right time. Architects the entire context pipeline: what data feeds into the model, how it is structured, what tools the model can call, what memory it retains.

Builds applications on top of LLMs. RAG pipelines, agent orchestration, tool calling, prompt engineering, eval frameworks. Works above the model layer -- composes models into products.

Deploys AI systems at customer sites. Part engineer, part consultant, part sales engineer. Makes the product work in the real world with messy data and ambiguous requirements.

Ships pixel-perfect, interaction-rich UI without handoff. Owns both the Figma file and the React component. Leads projects from design initiative, not engineering tickets.

Ships full products using AI coding tools without deep expertise in any single stack. Breaks down problems into tasks AI can execute, orchestrates multiple agents in parallel, and iterates rapidly. Measured by output shipped, not lines written.