Matt Busel

One-person studio (Tensorust). 8 iPhone apps live, 69 Rust crates, O'BLOCK on Steam. I write up builds with the real code and real numbers. tensorust.com

New York Joined on Oct 9, 2026 https://tensorust.com/articles/?utm_source=devto

Work

AI Engineer @ Tensorust

Matt Busel

Oct 18

App Store review time: 64 real decisions, median 73 hours

#ios #swift #mobile #appstore

15 min read

Want to connect with Matt Busel?

Create an account to connect with Matt Busel. You can also sign in below to proceed if you already have an account.

Create Account

Already have an account? Sign in

Matt Busel

Oct 18

LLM proxy benchmark: 200 duplicate calls became 1 (and a 30 s bug)

#rust #llm #opensource #performance

17 min read

Matt Busel

Oct 17

LLM JSON output: one setting dropped valid JSON from 75% to 4%

#ai #llm #javascript #webdev

13 min read

Matt Busel

Oct 17

iPhone app development cost: 20 real apps, $99 in fees, 8 approved

#ios #swift #mobile #startup

14 min read

Matt Busel

Oct 17

Landing page grader tested on 2,077 Show HN pages: correlation 0.00

#showdev #webdev #javascript #seo

14 min read

Matt Busel

Oct 17

crates.io download count: 80% of ours arrived with a new version

#showdev #rust #opensource #datascience

12 min read

Matt Busel

Oct 17

Run an LLM in the browser: WebGPU did 63 tokens/s, WASM did 1.6

#webdev #ai #llm #javascript

15 min read

Matt Busel

Oct 17

Bevy instancing: drawing 639,000 units with one draw call per mesh

#rust #gamedev #bevy #performance

13 min read

Matt Busel

Oct 17

Build an iOS app on Windows: how 21 apps shipped with no Mac

#ios #swift #githubactions #mobile

11 min read

Matt Busel

Oct 16

We cut the AI replies. Daily impressions fell from 447,575 to 5,686

#ai #agents #marketing #buildinpublic

6 min read

Matt Busel

Oct 14

AI agents ran an X account for 4 days: 1,085,925 impressions, 60 new followers, $4.99

#ai #agents #marketing #buildinpublic

10 min read

Matt Busel

Oct 13

Day 4: an AI agent's joke hit 306K views. It got 14 followers.

#discuss #ai #agents #marketing

8 min read

Matt Busel

Oct 12

Day 3: my AI agents got 500K views, 3 checkouts and $0. Here's what they changed.

#discuss #ai #agents #marketing

7 min read

Matt Busel

Oct 12

I let AI agents run my socials. Day 2: 220K impressions on an 87-follower account. Then Reddit banned them.

#discuss #ai #agents #marketing

6 min read

Matt Busel

Oct 11

Your coding agent's token just failed mid-task. Does it stop, or go find another key?

#discuss #ai #security #programming

7 min read

Matt Busel

Oct 11

AI writes code faster than anyone can review it. What do you actually check?

#discuss #ai #codereview #programming

6 min read

Matt Busel

Oct 10

I tried to make Claude Code cheat on tests. Here's when it did.

#ai #claude #testing #opensource

3 min read

Matt Busel

Oct 10

Coding agents keep "fixing" tests by editing them, so I made it impossible

#ai #claude #testing #opensource

2 min read

Matt Busel

Oct 10

My Windows wallpaper solves the geodesic equation, live

#showdev #rust #math #graphics

2 min read

Matt Busel

Oct 10

I ran 20 AI coding agents on one PC. The bottleneck was the compiler.

#ai #rust #productivity #programming

2 min read

Matt Busel

Oct 10

My game has one mechanic, so I taught the NPCs to lie

#showdev #gamedev #rust #indiedev

2 min read

Matt Busel

Oct 9

Making two cartoon eyes feel alive: springs, saccades and blinks in SwiftUI

#showdev #swift #ios #animation

3 min read

Matt Busel

Oct 9

Adding Chinese, Japanese and Korean to my Rust game engine without touching 39,000 lines

#showdev #rust #gamedev #i18n

4 min read

Matt Busel

Oct 9

I stopped paying twice for the same LLM answer (and why embeddings alone can't dedupe prompts)

#rust #llm #ai #opensource

3 min read

DEV Community

Matt Busel

Badges

1 Week Community Wellness Streak

Writing Debut

Skills/Languages

Currently hacking on

Available for

App Store review time: 64 real decisions, median 73 hours

Want to connect with Matt Busel?

LLM proxy benchmark: 200 duplicate calls became 1 (and a 30 s bug)

LLM JSON output: one setting dropped valid JSON from 75% to 4%

iPhone app development cost: 20 real apps, $99 in fees, 8 approved

Landing page grader tested on 2,077 Show HN pages: correlation 0.00

crates.io download count: 80% of ours arrived with a new version

Run an LLM in the browser: WebGPU did 63 tokens/s, WASM did 1.6

Bevy instancing: drawing 639,000 units with one draw call per mesh

Build an iOS app on Windows: how 21 apps shipped with no Mac

We cut the AI replies. Daily impressions fell from 447,575 to 5,686

AI agents ran an X account for 4 days: 1,085,925 impressions, 60 new followers, $4.99

Day 4: an AI agent's joke hit 306K views. It got 14 followers.

Day 3: my AI agents got 500K views, 3 checkouts and $0. Here's what they changed.

I let AI agents run my socials. Day 2: 220K impressions on an 87-follower account. Then Reddit banned them.

Your coding agent's token just failed mid-task. Does it stop, or go find another key?

AI writes code faster than anyone can review it. What do you actually check?

I tried to make Claude Code cheat on tests. Here's when it did.

Coding agents keep "fixing" tests by editing them, so I made it impossible

My Windows wallpaper solves the geodesic equation, live

I ran 20 AI coding agents on one PC. The bottleneck was the compiler.

My game has one mechanic, so I taught the NPCs to lie

Making two cartoon eyes feel alive: springs, saccades and blinks in SwiftUI

Adding Chinese, Japanese and Korean to my Rust game engine without touching 39,000 lines

I stopped paying twice for the same LLM answer (and why embeddings alone can't dedupe prompts)