Tim FinchAI engineer, systems builder

Writing

Product notes, systems thinking, and the working record behind what gets built.

A restrained archive of writing on product engineering, interfaces, and applied AI, alongside an in-progress build thread for Hoxa.

Featured Series

Hoxa Build Thread

A build thread documenting the product, design, training system, and architecture decisions behind Hoxa, an adaptive fitness app for beginner to intermediate users.

Seven posts / Product, design, architecture, and process

  1. Part 1 / April 12, 2026

    Why I'm Building Hoxa

    Hoxa starts from a simple observation: many fitness products lose people long before training becomes useful, because the product asks them to perform confidence they do not yet have.

  2. Part 2 / April 14, 2026

    Designing a Calmer Fitness App

    The visual and interaction tone of Hoxa has to lower friction without flattening ambition. Calm should come from clear priorities, not from stripping the interface until it feels generic.

  3. Part 3 / April 16, 2026

    Structuring the Hoxa MVP

    The MVP should be broad enough to express the product thesis, but narrow enough that every feature contributes directly to a credible training loop.

  4. Part 4 / April 18, 2026

    Adaptive Training Engine Notes

    The training engine should adapt in visible, defensible ways. If users cannot understand why the week changed, the system will feel clever before it feels trustworthy.

  5. Part 5 / April 19, 2026

    Hoxa Architecture Overview

    The Hoxa architecture should support adaptation, accountability, and future integrations without turning the product into an opaque tangle of services too early.

  6. Part 6 / April 21, 2026

    What AI Should and Should Not Do in Fitness

    AI can make fitness software more responsive and more understandable, but only if its role stays narrow enough to be earned and explained.

  7. Part 7 / April 23, 2026

    Building Hoxa in Public

    Writing the build thread is part of the product work. It sharpens decisions, exposes weak assumptions, and leaves a clearer record of how the system is taking shape.

Archive

Standalone notes on systems, interfaces, and applied AI.