Hilink
Explore
Services
product design, web design, frontend, backend
Year
2024 - Present
lUg2pdjrRyGVJUhotoy7Odjfns
What is Hilink?
Hilink is an AI‑assisted LinkedIn post generator and scheduler. Draft posts with an integrated assistant, attach media, and schedule posts that include a prepared comment thread to publish automatically at the time you choose.
Hilink was born from a limitation: LinkedIn’s native scheduler can’t attach comments to scheduled posts. Hilink bridges that gap while keeping the workflow fast, reliable, and creator‑friendly.
Key Features
- AI writing assistant for outlines, tone, and hashtag suggestions
- Schedule posts with prepared comment threads to publish together
- Direct publishing to LinkedIn (no copy‑paste)
- Rich media: images, videos (with custom thumbnail), PDFs, multi‑image carousels
- Offline drafts: persist post, attached media, and up to 2 comments locally so a refresh won’t lose your work
- Live preview (desktop and mobile) before scheduling
- Calendar and queue views with conflict and time‑zone awareness
Technical Highlights
- Next.js 15 + React 19, TailwindCSS
- Background scheduling with Trigger.dev and reliable retries
- Data layer with Drizzle ORM + PostgreSQL (Neon), caching with Redis
- Auth via Better Auth, payments/limits via Polar
- Media handling via Cloudinary; LinkedIn integration through official API client
- Schema validation with Zod; forms with React Hook Form; state via Zustand; data fetching with TanStack Query
- Observability and analytics with PostHog; deployed on Vercel
The Challenge
LinkedIn’s native scheduler doesn’t allow attaching comments at schedule time. We needed a reliable way to publish the post and its comment thread atomically at the scheduled time while handling LinkedIn media uploads, API rate limits, and transient failures.
The Solution
Hilink stores the post, media, and comment thread as a single scheduled job. At run time, a Trigger.dev workflow publishes the post via the LinkedIn API, uploads/attaches media (images, videos with thumbnail, PDFs), and immediately posts the comments, preserving order. Client‑side offline storage keeps drafts (including attachments and up to three prepared comments) safe between sessions. Idempotent operations, retries with backoff, and defensive validation ensure consistent publishing under API constraints.
Project Showcase
