Startup Stories - Mixergy
Episode #2295: Ryan Carson: How AI Does My Marketing For Me
Date: January 26, 2026
Host: Andrew Warner
Guest: Ryan Carson (Founder, Untangle)

Episode Overview

In this insightful episode, serial entrepreneur Ryan Carson shares with Andrew Warner how he uses AI to automate and supercharge his startup’s marketing efforts. Centered around his new company, Untangle (an AI-driven divorce platform), Ryan explains his journey from traditional email drips to deeply personalized campaigns, powered and orchestrated almost entirely by AI agents and automation tools. The conversation offers a practical, step-by-step walkthrough—complete with technical details, real-world anecdotes, and strategic advice—showing how startups can leverage AI to acquire and engage customers more effectively and efficiently than ever before.

Key Discussion Points & Insights

1. The Old Way vs. The AI Way in Email Marketing

• Traditional approach: Pre-written, templated “drip” campaigns with minor personalization (like a customer’s first name).
• The AI opportunity: “Why isn’t every email absolutely customized for the customer, so it’s completely personal? Like, we can do this now.” — Ryan Carson [00:55]
• Ryan built a system where AI writes every email drip personalized to the individual user, using any contextual information available.

2. How the System Works – Technical Walkthrough

Ryan describes his end-to-end process, built with modern AI platforms and automation tools:

Step-by-Step Workflow [03:59–12:47]

• Lead Entry: User interacts with a lead magnet (e.g., a child support calculator).
• Context Gathering: Store all available data about the user; enrich with info gathered from sources like Parallel AI, LinkedIn, and the X API.
  “You can actually use tools like Parallel AI to go out and scrape the Internet to find more information. You can get a lot of context on that customer.” — Ryan Carson [05:56]
• Campaign Creation: Map collected data to a custom email drip campaign table.
• Email Generation: For every scheduled email, an LLM (Opus 4.5 or GPT-5.2) gets a system prompt, user data, interaction history, and rules.
  • Include strict guidance to prevent AI from “hallucinating” features.
  • “You do not want the LLM hallucinating features for your app. That’s always terrible.” — Ryan [07:05]
• Delivery: Use tools like Resend (preferred for its ease of use and SEO/Answer Engine Optimization), Sendgrid, etc.
• Engagement Tracking: Monitor clicks and bounces, not opens, to assess engagement.
• Drip/Nurture Loop: After sequence completion, move contact onto a long-term, context-aware monthly nurture sequence.

3. Creating Real Results with AI-Driven Insights

• Example: The AI, monitoring user engagement, spotted “five people that have clicked three different times on your email drips.”
  Ryan had the AI send a personal email offering a call, instantly resulting in a new customer:

  • “I immediately got a reply from one of those customers … and she said, I would love to meet. ... In the end, she’s like, I’m gonna sign up for the free trial.” — Ryan [02:51]

• The AI not only handles communication but also recommends actionable next steps every day.

  • “It generates an email with recommendations on what is the one thing I should do to improve my marketing funnel today.” — Ryan [01:51]

4. The Power of Specialized Vertical AI Agents

• Untangle’s AI agent “Grace” is trained and tooled specifically for the divorce niche (e.g., statute search, guideline lookup).
• Grace writes both legal content and customer emails.
• Ryan’s core advice: “You’re going to have an agent that has specific skills. That should be the agent that writes your content, that should be the agent that writes your emails. Like this is all the same entity.” — Ryan Carson [13:19]

5. Automation Tools: Amp, Ghosty, Zapier, and More

• Building the Stack:
  • Amp: Advanced coding agent; Ryan uses it (within Ghosty) to build and extend his stack (NextJS, Typescript-based).
  • “All I did is open up Amp and said, I want to build an email drip.” — Ryan [16:30]
• Task Handoff: Ryan runs multiple Amp agents (through different Ghosty windows) simultaneously—delegating several engineering tasks at once.
• Integration Example: Zapier is used for connecting disparate tools and automating data flows.
  • Andrew: “They actually will have a connector into Claude that connects to all their other tools. … Zapier just pulled it together and created an email in Gmail for me.” [15:06–15:46]

6. Building & Managing with AI

• Ryan describes his workflow of ideation (using Fizzy for backlog), breaking down requirements, and using the “RALPH loop” to implement features via AI agents.
• “We live in a world now where effectively code is free, it's at least very cheap … so what we build is now becoming more important than how fast we build it.” — Ryan [21:24]

7. Daily Marketing Optimization Loop

• Ryan’s most valuable loop: An AI-generated daily email with one or two concrete marketing action items.
  • “Every day just go to Amp … and say, I want an email every day with one to two key actions I should take to acquire more customers at our desired CAC.” [24:05]
• Feeds data from lead magnets, conversion steps, and drop-offs into the AI for diagnosis and incremental improvement.
• “All you're doing is increasing your batting average … it's more likely that I'll see a good suggestion if I'm getting a smart suggestion once a day.” — Ryan [31:23]

8. Insights, Iteration, and Human Judgment

• The system surfaces valuable but often basic actions—things a human marketer might miss or neglect.
  • Ryan: “All that Amp is doing for me is really reminding me of fairly basic, obvious things that I need to do that I haven’t done yet. And that's actually very valuable.” [30:39]
• Ryan stresses the importance of context: “The agent needs context to give you valuable data.” [26:00]
• Ideas for lead magnets (like calculators) can, in fact, be generated by the AI—Ryan’s own was.
  • “I'd say, Amp, we need to capture leads. What should we do? … We could calculate child support, … create a roadmap for divorce. … It turns out calculating child support is extremely complex, and you have to have all the right tools, but we've built those.” [27:36]

9. The Future: Full Automation of Marketing Implementation

• Ryan aspires to have his AI not just suggest but actually implement its marketing recommendations—submitting PRs for real feature changes overnight.
  • “The next step is not just to get the email, but actually have Amp implement the suggestion. And then I wake up to a PR that I review.” [28:40]

10. Advice to Founders: Start Simple, Iterate

• Don’t overthink AI’s role—start with custom LLM emails, then evolve.
  • “Most people aren't even doing what I showed you … Start there and then iterate and then eventually after a year, you'll be this machine. But just take step one.” [33:05]

Notable Quotes & Memorable Moments

• On the promise of AI in marketing:
  “Email automation, AI—you know, all this is beginning to work. It was really fun.” — Ryan Carson [02:51]

• On context as the key to AI value:
  “You need context. The agent needs context to give you valuable data.” — Ryan [26:00]

• On AI’s marketing insight:
  “It's not rocket science, but it's an actionable, valuable insight.” — Ryan [31:23]

• On the founder’s new challenge:
  “What we build is now becoming more important than how fast we build it. … Are we building the right things quickly enough? So the way I look at it is…I'm just trying to be vulnerable in the sense that everybody's dealing with this question of are we building the right things?” — Ryan [21:24]

• On future trends:
  “Somebody should be watching this and saying … I think what we could do is create really totally customized email. … Not only that, we start to enhance it with your own data and that becomes proprietary internal information.” — Andrew Warner [34:04]

• On letting AI implement decisions:
  “The next step is … have Amp implement the suggestion. And then I wake up to a PR that I review where it’s actually done the thing that it suggested.” — Ryan [28:40]

• On AI-assisted ideation:
  “Go talk to your agent and then iterate and then measure, and then iterate and measure as fast as you can.” — Ryan [27:32]

Important Timestamps

• 00:55 — Ryan on why every email should be customized, the birth of his AI-driven system.
• 01:51 — Example of AI-generated email insight leading to a new customer.
• 03:59–12:47 — Ryan’s end-to-end technical demo of his AI email marketing workflow.
• 13:19 — Specialization of vertical AI agents (“Grace”), and their expanding roles.
• 14:26–16:14 — Zapier’s role in Ryan and Andrew’s automation stacks.
• 16:30–21:24 — Building infrastructure with Amp, AI coding agents, and Ryan’s workflow philosophy.
• 24:05–31:23 — How Ryan gets daily actionable marketing advice from AI and applies it for growth.
• 27:36 — Using AI for marketing ideation and iteration.
• 28:40 — Vision of AI agents implementing their own recommendations.
• 31:23 — AI surfacing basic (yet high-leverage) growth opportunities.
• 33:05–34:04 — Founders are urged to start simple, iterate, and eventually outpace competitors with sophisticated AI automation.

Takeaways for Founders & Marketers

• AI-driven marketing is here—and works best when fully personalized: Even basic, daily suggestions from AI can profoundly improve customer acquisition and engagement.
• Provide context: The more relevant data you feed your AI, the more actionable and effective its suggestions.
• Start small: Even simple improvements via AI quickly compound; don’t wait for “perfect,” start automating and personalizing now.
• Leverage specialized agents: Strongly tool your AI for your industry/vertical, and let it produce content, campaigns, and more.
• Aim for the “autonomous marketer”: Expect a future where AI not only tells you what to do, but does it for you.

Summary in a Sentence

Ryan Carson demonstrates how founders can use off-the-shelf AI tools and coding agents to build highly effective, fully personalized marketing systems—boosting customer acquisition with daily, actionable insights, without expensive hires or heavy manual work.