AI Workflow Design

Nate Herk's Three AI Workflows: Video, Decisions and Instagram

Can an agent edit videos like a human team, challenge a business decision and run a social account? In Matt Wolfe's Show Me Your Workflow episode, Nate Herk shows working versions of all three. The transferable idea is not a single magic prompt. It is a loop: study approved work, write down the rules, produce a draft, inspect the result and turn specific feedback into the next rule. Nate's results are demonstrations of his own systems, not a benchmark showing every creator will get the same output.

The useful pattern

Pick one recurring job with examples of what good looks like. Give the agent a narrow skill and the right tools, then keep a human review point where taste, facts, rights or publishing authority matter. Start with one draft and measure the repair work before adding more automation.

Watch Matt Wolfe and Nate Herk

What is shown: Nate walks through his video editor, the Roast Room and a social account. I checked the linked public projects and vendor documentation on 29 October 2026; I did not run Nate's private pipeline, verify its analytics or compare its output with his human editors independently.

1. Teach an Agent Your Editing Style

Nate began with 27 human-edited reels that had performed well for his account. He studied their pacing, cuts, scene types, captions, music and sound, then put the recurring patterns into skills. In the interview he compares an original edit with an agent-made one and says the goal is to preserve his team's recognizable style. His public editing playbook describes the same feedback loop: a rejected caption, awkward cut or weak sound becomes a rule or check for the next project.

  1. Transcribe and map the recording. Word-level timing helps place cuts and captions. Nate uses ElevenLabs Scribe; its documentation confirms timestamped transcription. The kit also allows an existing transcript or a local transcription route.
  2. Plan the cut before the polish. Preserve the speaker's meaning, choose the clean takes, identify visual beats and show proposed mistake cuts for review.
  3. Build and inspect a draft. Nate uses Claude Code with HyperFrames to assemble motion, captions and media. His agent inspects frames and renders, while he still judges pacing and sound.
  4. Save the correction. A one-off note fixes one video; a tested rule, reference or script check can improve the next one.

The public Herk Edit Kit contains skills, scripts, examples and a synthetic demo. It is not a ready-made license to reuse Nate's brand assets: the repo's license note excludes the AI Automation Society branding. Third-party transcription, generated footage, music and model use can add costs. The first useful test is one recording you own, one style brief and a reviewed draft.

2. Use a Panel to Challenge a Decision

Nate's Roast Room gives simulated advisors different questions to defend: product demand, sales, marketing, cash, operations and founder time. Claude Code acts as chair. The agents argue across rounds and produce a scored report; the animated table is a display of those files, not a separate reasoning engine. Matt tests it on the future of his sponsor-funded newsletter.

The result is useful as structured opposition, not as a board vote. Ask it to compare real options, state the numbers and constraints, surface the strongest objection and propose a reversible test. A panel made from the same underlying model and source material does not become seven independent experts merely because it has seven portraits. The repo explicitly says these are simulations, not endorsements or statements from the named people. It also notes that a full council uses 21 sub-agent runs; the quick mode is more economical for small decisions.

For a high-impact financial, legal or hiring call, verify the underlying evidence and make the decision with qualified humans. If the panel's biggest contribution is a question you had not asked, it has already earned its place.

3. Research, Make and Publish Posts

The third demo is Nate's Instagram account. He says a scheduled run gathers candidate stories, checks sources and reactions, creates branded carousels or clips, writes captions and publishes through Blotato. A Google Sheet records published topics to reduce repeats. Blotato's own site documents publishing, scheduling, DMs and analytics through an agent-facing API and MCP integration. This is a paid distribution service, not one of Nate's open-source repos.

The important boundary in the video is easy to miss: Nate still adds interesting posts himself, rejects weak output and asks for revisions. He says the performance-feedback stage is only partly set up, not a mature self-improving loop. His account's early follower and time-spent numbers are his observations, not a forecast for a new account.

A practical editorial gate sits between research and publishing. Check the original source, date, claim, context and reuse rights. A link to a public video does not grant permission to repost its footage; credit alone is not a license. Preview the visual on a phone, approve the caption and final asset, then authorize the post. Recheck time-sensitive claims immediately before publication, as Nate demonstrates. An agent can do the preparation; accountability for what appears under your name stays with you.

Which Workflow Should You Try First?

Your bottleneckFirst testWhat would count as progress
You edit the same video format every weekGive the editor one owned recording and two approved examples. Ask for a draft, not a publish-ready final.Less repair time, preserved meaning and an edit the owner would approve.
You circle a decision without testing assumptionsRun a quick Roast Room round on two real options and a fixed budget.A sharper question or cheaper experiment, not a confident-sounding verdict.
You have ideas but cannot publish consistentlyResearch and draft one sourced post, then approve it manually before connecting a scheduler.Faster production without factual, rights or brand errors.

Do not automate the full account to discover that the research or editorial judgment was the weak step. Prove one unit of work, then widen the loop.

Repos and Tools by Role

ResourceRole in the workflowBuilder note
Herk Edit KitPublic skills and scripts for transcript-led video editing.Read its setup and license first; replace Nate's brand assets and review every cut and sound choice.
Roast RoomPublic decision-debate skill and optional local visual room.Simulated viewpoints are not those people's advice; full runs consume substantial model usage.
HyperFrames and ScribeVideo rendering and word-timed transcription used in Nate's edit pipeline.Rendering needs a local toolchain; the hosted transcription API and optional media generation may cost money.
BlotatoPublishing, scheduling and DM/analytics integration for the social demo.Separate creation from posting permission. Nate's exact Instagram orchestration was shown in the interview, not released as a verified repo here.

A One-Video Pilot Prompt

There is a credible service idea here: help one client turn repeatable recordings into a consistent, reviewed content format. The following is an original editorial prompt for testing that offer, not Nate's prompt or a workflow we have run.

Business pilot

One recording, one approved edit

Test quality and repair time before promising scale.

Ready to copy

Video Chapters

Editing in Nate's style

TimeTopic
00:00Matt Wolfe introduces Nate's three workflows
01:03Study proven edits and extract style rules
06:33Claude Code, transcription, media and rendering

The decision panel

TimeTopic
13:50How the Roast Room is built
20:37A live newsletter-strategy debate

The social publishing loop

TimeTopic
26:49Research, create and publish Instagram content
36:24Where to start with Nate's workflows

Sources and Limits

Checked 29 October 2026. The Matt Wolfe interview and supplied transcript support Nate's demonstrations, his 27-reel analysis, his account observations and the parts of the social loop he says are unfinished. Nate's Edit Kit and Roast Room document the public projects. HyperFrames, ElevenLabs and Blotato document supporting tools. We did not install the projects, reproduce the edit comparison, audit the account, or find a public repo for Nate's exact Instagram pipeline. Check current setup, pricing, permissions and media rights before using them.

Common questions

Can I download Nate Herk's video-editing workflow?
Yes. Herk Edit Kit is public on GitHub and includes skills, scripts, examples and setup guidance. It is a starting point for your own style, not a promise of a finished edit without review.
Does the Roast Room contain advice from the real people it depicts?
No. The seats are AI simulations based on cited public material. They are not the people and do not endorse the project. Treat the output as a way to test assumptions, not as professional advice or a final vote.
Is Nate's Instagram system completely automatic?
Nate describes a mostly automated account, but says he still supplies ideas, reviews weak posts and has not fully completed the performance-feedback loop. The transcript does not provide a public repo for this exact social pipeline.
Share
X LinkedIn Reddit
Build Yours

Want a system
like this one?

Book a free 30-minute call. We map your situation, identify the highest-impact automation, and figure out if we are a fit.

Book Free 30-min Call