Hesya

Existing tools for attention and screen time split into two failure modes: apps that pull people into unconscious digital escape when they’re already overwhelmed, and productivity tools that push control, guilt, and optimisation. Neither restores agency — one exploits distraction, the other polices it. Hesya was built to sit in between: a session protector for knowledge workers that helps them notice drift and return to what they meant to do, without becoming a system they depend on.

Ship a native iOS app, solo, that helps people stay oriented to a declared intent — recovering from distraction without shame or blocking — where success is explicitly measured by the tool being needed less over time, not more. Silence is a valid, intended outcome.

Fifteen interviews before a line of code. Not a survey. Long conversations with knowledge workers who already recognised the problem in themselves.

Five things came back repeatedly, and each one became a constraint:

“I know I’m drifting. I just can’t stop.”

Awareness isn’t the gap. Detection alone solves nothing — the product has to offer a way back, not a warning.

“I feel guilty afterward, but I couldn’t stop in the moment.”

Shame is already present. Anything that adds to it makes the product part of the problem it claims to solve.

“Focus apps became another thing to manage, not less.”

The product has to be abandonable. Success looks like needing it less.

“Too many choices. I freeze instead of starting.”

Three taps to start. Nothing pre-filled. One intention, not a task list.

“I opened my phone to scroll and lost an hour I didn’t have.”

The cue has to arrive during the drift, not in a summary afterwards.

I tested the copy as carefully as the interface. Every cue was checked against the words people actually used — in a product about attention, the tone of a prompt is the product. The same sentence can read as an invitation or as a scold.

• Designed the core loop — Declare → Start → Stay → Recover → Reflect — as the spine of the product, with supporting cues kept sparse by design
• Built drift detection using Apple’s Screen Time / DeviceActivity APIs: sessions schedule a threshold on user-marked “pull-away” apps; on trigger, a background monitor checks cue gates before surfacing a prompt
• Designed pacing cues as an opt-in, separate mechanism from drift recovery — silenced automatically if a session looks fragmented or is paused
• Designed the shame-free reflection loop and Journal / Weekly Pulse pattern views, shown only when the user asks to see them
• Built the full SwiftUI implementation solo — product philosophy, UX, on-device pattern recognition, and App Store release — with no external dependencies or data collection

Shipped solo, end to end: philosophy, discovery, UX, SwiftUI implementation, App Store release. No team, no external dependencies, no data collection.

A private beta ran before public release to validate the Declare → Stay → Recover → Reflect loop in daily use. The loop held. What changed was narrower: intent inputs needed restyling for Cyrillic, and the language around cues was tightened again after seeing how it read in the moment.

Live on the App Store. Free, no account, nothing to cancel.

Since release on 27 July 2026, through 9 September: 158 App Store impressions, 47 product page views, 11 first-time downloads. Small numbers, shown rather than omitted.