The Founder Who Wanted Proof, So He Built It

Legal tech founder, AI Edge for Working Professionals

A sunlit startup meeting room with a large monitor showing a connected workflow dashboard and charts

He runs a startup that builds tech solutions for legal teams, and he was already using Claude. What he lacked was structure. "It's largely been me just experimenting myself and then obviously, you know, connecting with my CTO and saying, hey, I want to do this, I'm struggling".

He also knew himself well. Before the first session, he asked for one thing above all. "I have a knack for kind of procrastinating a bit, so I want to be held extremely accountable".

There was a bigger reason, too. He believed in what AI could do for his company, but belief alone was not going to move his team. "a couple of people in my team are still skeptical. How much AI can do that, and I, at least in my mind, do believe it can, but I'm not able to prove that. So I need your help..."

He joined AI Edge for Working Professionals, Arthi's 1:1 mentorship, and the early sessions were about getting his hands dirty. He worked through Claude Cowork and Claude Code, built skills and scheduled tasks, and made his first app, a property rental search tool. Then came something closer to home: a household meal-planning app. "My family laughed at the idea, but I thought it's practical because everyone struggles with it at home." He deployed it through GitHub and Vercel.

The builds kept coming. A skill that tracks weekly LinkedIn signals. A LinkedIn profile search app. A resume agent running on local Ollama models. Along the way he explored agents, vector databases and ElevenLabs voice agents.

When the package wrapped up, about five weeks after his first session, he was warm and direct. "It's been great, you've been amazing. Thanks so much. If you need any reviews, please hit me up. I'll be happy to share whatever you need."

But the real turning point came about a week later, at a follow-up. He wanted to start by sharing his screen. "I will just share my screen and I just also want to show you what I have done. So as a teacher you can be happy."

What he showed was a company in motion. He had mapped the manual work in each department and set up agents across sales, customer success, finance and management.

"So, you know, I've been able to successfully implement a bunch of things. So firstly, just wanted to show you, you know, what you've taught me. I've been able to implement, you know, quite a bit, you know, in reality."

Agents now research prospects and prepare talking points before a demo. "I've automated my entire LinkedIn messaging, so if you see, this is all Claude, you know, sending messages to people for me and I've even now looked at AI voice agent." An agent for LinkedIn posting and his newsletter was due to be finished that week. Claude had turned his PDFs and LinkedIn into on-brand creatives and a feature video. "So if you see, this is like too good."

The voice agent idea came straight from an ElevenLabs demo Arthi had shown him in an earlier session. He built a customer success voice agent for check-ins and payment follow-ups, and he could already see where it fit. "but for check-ins and payment, I think it can work beautifully and help us. So, you know, these are things that have been great for me."

The founder who once took his questions to his CTO was now the one showing what was possible. Looking back at everything he had learned to do, he put it simply. "This was pretty cool that I was able to kind of pick up."

Want to build something like this? Explore Arthi's 1:1 programs.

Read more learner stories or browse testimonials.