HN Showcase
Back to browse
What if your personal website was ChatGPT?

What if your personal website was ChatGPT?

by mrbnprck·Jul 7, 2026·1 point·8 comments
Visit ProjectView on HN

AI Analysis

●●SolidCozyWizardry

Server-side Swift + HTMX with intentionally throttled Groq responses for natural chat feel.

Strengths
  • •Zero client-side JavaScript except markdown rendering and HTMX.
  • •Custom server-side Swift stack for static site generation and networking.
  • •100-200ms latency with deliberate throttling for conversational pacing.
Weaknesses
  • •Anti-crawler design hurts SEO and discoverability for a personal site.
  • •Novelty interface that most visitors would prefer as scannable static content.
Category
Design
Target Audience

Developers, designers building personal sites

Post Description

Hi HN,

I recently updated my website to have a touch of this new AI thing.

The initial idea was simple: what if visitors could explore a CV by chatting with it instead of scanning a traditional static website?

I'm quite happy how the final version turned out:

* It runs LLM inference via Groq at 100-200ms latency at super low cost. The token generation is intentionally throttled to make the interaction feel natural, otherwise you'd literally hit enter and the response is already there.

* Except for a markdown rendering and HTMX4 there's no client side javascript!

* It's build with a custom server-side swift stack so everything related to the HTMX4 static site generation and networking layer is pure swift.

* As a nice side effect, search engines/crawlers cannot collect my CV right away, everything is bound to interactivity with the chat interface.

Technically you can ask the chat anything you want about my professional work experience. Rate limits & guard rails apply :)

I'd love to hear your thoughts.

Similar Projects

Developer Tools●Mid

AccessLint – Static accessibility analysis for iOS/Swift

It reads your Swift sources and flags accessibility problems with exact line numbers, suggested fixes, and WCAG references — handy for catching issues in CI or locally via a Homebrew-installable CLI. Practical execution is clear (SwiftUI/UIKit rules, Xcode/CI output), but the rule set is small and several checks are behind a paywall, so it's a strong team tool rather than a category-redefining product.

Niche GemSolve My Problem
synctek
104mo ago
AI/ML●●●Banger

Static-allocation MLP inference in ANSI C using a 2-slot ring buffer

Two-slot ring buffer cuts MLP RAM usage to the practical lower bound on microcontrollers.

Niche GemWizardryBig Brain
xou
401mo ago