Engine AI

AI Coding Platform

  • Year

    2024-2025

  • Website

    Client / Freelance

  • My Role

    Lead Product Designer

Product

Objective

Engine AI is an AI platform built for experienced engineers who want architect-level control, not generic copilots.

As the first design hire, my role was to define the product UX, interaction model, and design system in a category with no established patterns, while shipping quickly enough to validate product–market fit.

Core tensions

  • Power users wanted raw control, not abstraction

  • New users needed guidance without dumbing things down

  • LLM tooling was powerful but mentally expensive to operate

  • Mobile traffic was high, but the product was desktop-only

Process

  • Phase 1: De-risking the UX through rapid MVP iteration

  • Phase 2: Establishing a scalable design system and interaction model

  • Phase 3: Evolving the product from MVP to the Bento architecture

Outcome

The redesigned MVP reduced core UX friction and enabled faster iteration across product and engineering.

Mobile designs were introduced for the first time, addressing a significant portion of the user base and improving early-funnel engagement. Qualitative feedback from the community (Discord) indicated improved clarity, confidence, and perceived control when working with the platform.

Longer-term, the project established the foundation for the Bento architecture; a design direction that balanced power and approachability and continues to guide the product’s evolution.

Standout Features

  • Wizard-based project creation for complex workflows

  • Shell mode for advanced users requiring direct LLM control

  • Bento-style layout for scalable, modular interaction

  • Mobile-first redesign for a previously desktop-only platform

  • Design system optimised for rapid iteration in an evolving AI product

Research + Personas

Research was conducted through internal workshops, affinity mapping, and continuous feedback loops with early users. The focus was on understanding mental models, tolerance for abstraction, and expectations around control when working with LLM-driven tools.

MarComs

Alongside the product, I contributed to brand and marketing assets to ensure consistency across the platform, website, and early acquisition touch points. This helped align product expectations with the underlying interaction philosophy.



Probably on my 4th espresso of the day