Life when credits run out
At the time you read this edition, I will be back in Hyderabad.
I am writing this from a small village in Dharamshala and I’m sad to leave this place.
Three weeks flew by in a snap of the finger, and I can’t wait to visit again.
Anyway… with so many memories, I also have beautiful things to look forward to at home. So, hello Hyderabad 👋
How I built an agent to outsource boring admin tasks
I run a community of marketers who geek out about using AI at work.
I loveeeee hosting sessions, asking good questions, and learning from each other.
But part of my job as a host also involves multiple admin tasks that aren’t my favourite, but important.
So I have built Pingy, our community.
It’s a Discord agent with some simple automations + AI, and it takes care of:
Onboarding: DMs new members about their first steps immediately after they join
Retention: Pings me 10-30 days before a member’s subscription ends (based on tier), so I can focus on retention by asking how I can improve the experience.
Analytics: I used this saas, Statbot. The free plan only gives access to two weeks of analytics, and I don’t own the data + no privacy. Now Pingy reads and analyses all member data privately and provides me with a weekly engagement report.
Newsletter: Our community is on Discord and it’s not a daily app for our members. So Pingy curates the best conversations of the week and sends an email every Monday to encourage inactive members to stay more active.
It costs nothing. I only paid for $20 for the Claude Pro plan. Used Claude code and Discord’s bot feature.
How Pingy reminds me of ending subscriptions
If I am being super honest, this is not a complex workflow to build. At the core, it’s a simple automation with AI.
But that’s how I measure AI applications in marketing.
I don’t really care about the tools I use, the complexity of the workflows, etc.
The job of agents is to solve a problem.
Pingy solves retention, engagement, and data privacy, so I consider this successful.
My friends ask me where to get started with AI in marketing.
I usually say find problems to solve. It’s easy to get excited by (and get distracted) by building apps, creating fancy dashboards, etc., which is important to understand what AI is capable of.
But once that is done, focus on problems, KPIs, and metrics that help the business.
In future versions of Pingy, I want to add a fully generated AI newsletter that shares Discord threads based on members' interests.
You might join the community to talk about organic marketing, and you’re not really interested in paid growth. Then Pingy should email you about organic growth conversations and shouldn’t bother you with irrelevant convos.
One day, I will share how this goes.
Marketing resources shared inside our community
How is the TMM community using AI?
This Modern Marketer created a thinking and content agent:
I am a founder and executive content thinking partner for CEOs and B2B brands.
In this AI era, I wanted to get better at business and thinking and more consistent with my own LinkedIn. So, I built two AI agents to help me do just that.
First is Rumi, an agent that helps me think better. It curates varied articles, then asks calibrated questions across six thinking dimensions like bias detection, metacognition, steelmanning and maintains a living record of my patterns in Notion. The goal is to help me read new interesting pieces of content each day and use it to inform my LinkedIn content, my pov, and my expertise.
Second, is Sophia, a LinkedIn content agent who helps me write consistently on LinkedIn on topics I care about. The agent reads my project files. Pulls from Rumi's Notion page and recent session logs to get a sense of my thinking patterns and my evolving pov. Proposes three content angles. Once I pick an angle it drafts the post and saves the approved post to Notion.
Submit your use cases and get featured in front of 1700+ readers
PS: This is my new newsletter. If you have feedback on what can be most useful to you, please let me know.
Have a lovely week. See you next Monday!