Brand Harvest

Custom software for an AI-native agency

Process design & technical implementationAI workflows & MCP integration

The context

Brand Harvest had the marketing expertise and client relationships. To operate as an AI-native agency, it needed software built around how its team actually delivers content. I built Contentool to connect that work in one system, from turning client interviews into ideas to checking content quality and managing distribution.

What I took on

Translate the agency’s content processes into a custom software system.
Connect client recordings, insights, topic planning, and content delivery.
Build AI workflows that check content quality and handle distribution automatically.
Expose the system through MCP so the team can operate it from its preferred AI agents.

The work

Contentool brings the agency’s content operations into one workspace. AI processes handle quality checks and distribution without a person having to run each step. The team can manage clients, work with material from interviews, plan topics, and review activity in the interface. Through MCP, it can also control the system from its preferred AI agents, using the same workflows beyond the app itself.

From client conversations to content

Contentool / Interview insights

Client interviews contain the experiences and opinions that make personal brand content useful. Contentool brings the recording, transcript, and extracted insights together, so the team can turn a specific idea into a post while keeping its original context close at hand.

Contentool showing extracted interview insights alongside the transcript and audio player — Extracted ideas connect back to the client’s words and recording.

A shared plan for every client

Contentool / Topic planning

Topic pillars give the team a structured view of what each client should be talking about. Coverage and linked posts make it easier to see which themes have been explored and where there is still room for new content.

Contentool topic pillars with coverage indicators and content topics grouped by theme — Client topics and coverage in one planning view.

Client activity in one place

Contentool / Client overview

The client overview brings recordings, posts, connected accounts, and LinkedIn connection history together. It gives the agency a single place to check what is happening across a client’s content work.

Contentool client dashboard with recording and post counts, a connected account, and LinkedIn connection history — A client workspace combining content activity and account tracking.

Work through an AI agent

Contentool / AI agent & MCP

The built-in AI agent lets the team ask questions about its work and get answers from the system. MCP extends that access to the agency’s preferred AI agents, so the team can control Contentool and its workflows from the tools it already uses.

Contentool’s built-in AI agent answering a question about client feedback and response times — The built-in agent answering operational questions. MCP also makes the system accessible to external AI agents.

An AI-native agency, in practice

Contentool automates 80% of the agency’s previously manual content work and helps it deliver better quality for clients. Quality checks and distribution run through automated workflows, while the team can oversee and control the system through the app or its preferred AI agents. My contribution was to turn the agency’s way of working into software it can use every day.