Take notes from Jayant Kumar and Vincent Oktavian on what’s worth automating in our latest session
We are living in the era where AI has quietly become our daily routine at work. Tasks that used to consume hours to finish, now take a fraction of time. ChatGPT and Claude, for instance, have become the go-to-tools for speeding things up.
But popular doesn’t always mean understood. Plenty of founders and business owners still aren’t sure what’s actually happening with the new technology, and the questions that follow are very practical ones. If your business is stuck in a problem and your team can’t solve, is AI the leap worth taking? Where should we begin?
In the midst of AI evolution, Foundry Collective is back with “AI-tomate Your Business” session last week, brought in two founders who know their way around using AI in a real business: Jayant Kumar (Founder of Emp0) and Vincent Oktavian (Founder & Operator of F&B business).
Here are some key takeaways from the conversation:
A peer, not replacement
Vincent opened the morning session with a question that most owners have faced at some point: what do you do when problems show up, but you can’t fix it yourself? The reasons may vary. Sometimes the budget isn’t there or the team simply lacks the skills the job demands.
As a founder running F&B outlets, he figured that AI can help close the gap in certain areas. He followed with what’s worth automating, and he narrowed it down to three: tasks that are repetitive, tasks with clear rules, and tasks where the data already exists (in email, Excel, and WhatsApp).
WhatsApp replies are one example, especially when it comes to handling hundreds of customer chats. After using an AI agent, he experienced an increase of replies sent within 15 minutes to 89% and customer chat volume grew by 70%.
Despite the results, Vincent was realistic about where this is headed to. Automating work with AI doesn’t mean one-person company has to be the end goal. In his view, “AI works best as a peer rather than a replacement, there to fill what’s missing, whether that’s a skill the team doesn’t have yet or productivity still running low.”
What AI offers, consistency and pace
When Jayant started building a channel about content automation with AI, there are three layers he emphasized: chat use cases, systems, and agents. Marketing and sales sit at the heart of every business—both usually run on content—which is also where many founders get stuck with resources they have.
Making one post by hand can take about 30 minutes or more, that’s where AI changes the process, and not just the speed. What he learnt from his tools is that comparing an AI-run page with a human-made: views climbed significantly. He pointed out that consistency and pace are exactly what AI could offer, and what people often struggle to keep up.
AI, still, doesn’t replace knowing your business, he stressed. Owners must understand which channel fits their product and market, and where revenue actually comes from. He gave one example from his work where YouTube lifted views higher than Instagram and LinkedIn, the platforms you’d usually expect to do the job.
Jayant was also candid about where AI can get things wrong. AI is basically trained to handle a wide range of tasks, including identifying objects, but there’s plenty of room for errors. As he put it, “AI is smart, but not smart enough yet to understand and identify the right image in an Instagram post, so be careful with source selection”.
So, is AI worth the leap of faith? The takeaway from both founders is a practical one. You can do the first step with the small or repetitive task and learn from what comes back.
***
