All projects

LinkedIn Intelligence

2026

Content Engine

Content strategyDraftingScheduling

Overview

An automated content pipeline for LinkedIn. It mines the topics and conversations an audience actually cares about, drafts genuinely useful posts — teardowns, breakdowns and lessons rather than announcements — and schedules them at a native cadence. A human reviews, adds the real engagement and hits post.

Role

Built the pipeline end to end — topic mining, drafting, per-platform re-angling, scheduling, and a tracking loop that logs which hooks and formats win and feeds that back into the next cycle.

The pipeline

Five stages — mine high-intent topics, draft, re-angle natively for the platform, schedule at a safe cadence, and track what wins. The only manual steps left are reviewing, the genuine replies, and hitting post.

Native, not copy-paste

Every piece is rewritten for the culture of the platform instead of one post blasted everywhere, so the account grows instead of getting flagged.

Highlights

Value-first drafting

Useful content, not ads or announcements.

Tracking loop

Winning formats and hooks feed the next cycle.

Team-of-two cadence

Runs a channel at the pace a larger team usually would.

System architecture

TopicsGenerateDraftsQueuePublishLogsFeedbackTopic MiningCommunity signalDraftingValue firstRe-anglingPlatform nativeSchedulerNative cadenceLLMDraft modelTrackingFeedback loopPlatform APIPublishing
Built by meExternal / third-party