AI Blueprint
All builds

IT Services

An IT company’s agency was posting content so boring it got literally zero engagement. The owner knew LinkedIn mattered but never had time to do it himself.

Built an automated content pipeline that finds relevant industry conversations, scores them against the brand, and delivers ready-to-post LinkedIn content every morning. The owner spends 5 minutes a day instead of an hour he never had.

5 min/day

to review and post

3 posts

delivered every morning, pick one

N8NClaudeGeminiRedditRSS FeedsObsidian
01

The Problem

The owner of an IT services company knew LinkedIn mattered for visibility and lead generation. But the agency handling their content was producing bland, generic posts that got zero engagement — not low engagement, literally zero. The content didn’t reflect the business, didn’t speak to their audience, and didn’t give anyone a reason to stop scrolling.

02

Before

An agency posted templated content a few times a week. No industry relevance, no brand voice, no point of view. The owner occasionally thought about writing something himself but never found the time. LinkedIn was a box being checked, not a channel producing results.

03

What I Built

A content pipeline that runs on autopilot. It scrapes relevant subreddits and RSS feeds for real industry conversations and trending topics. A scoring system evaluates each piece against the business’s brand context — who they are, who they serve, what they care about. High-scoring insights get summarized and stored.

As part of this build, I developed a brand voice document that captures how the business actually sounds — tone, language, point of view. That document doesn’t just power LinkedIn. It’s a foundation the business can use for anything they write going forward.

Every morning, an AI agent reviews the top-rated content — going back to the original source URLs to pull full context on what was flagged and why. It then researches the claims and information in each piece to make sure nothing gets published that isn’t grounded in reality. From there it writes three LinkedIn posts using the brand voice guidelines, runs them through a humanizing layer that still respects those guidelines, and delivers everything as a morning brief in Obsidian with image suggestions. Pick one, tweak if needed, post.

The agent can also be custom-queried with flags to produce research-backed content on specific topics on demand.

04

After

The owner opens a note every morning with three ready-to-post pieces that actually sound like him and reference real industry conversations happening that week. Five minutes to pick one, approve it, and post. No more staring at a blank screen, no more paying an agency for content nobody reads.

05

Results

The agency content that got zero engagement was replaced by a system that produces relevant, on-brand posts daily. The owner went from “I should post on LinkedIn” to actually posting consistently — 5 minutes a day instead of the hour he never had. Engagement has increased and followers are compounding daily. The system has been running for 6 months and stays current because it’s sourcing from live conversations, not recycling the same talking points. The brand voice document developed for this build now serves as the foundation for all of the business’s written communication.

Let’s sketch out where AI fits in your business.

30 minutes. No pitch deck. No pressure. Just a conversation about your business and whether AI makes sense right now.

or send a message