How a Private Chef Startup Went All In on AI Agents

1. SMSiddhi Mittal0:02

   Hey everyone, I'm Siddhi, the founder of Yhangry. Gonna cover three business use cases that we're using Yhangry for. There are like a million. Okay, engineering. So this was when I was on math leave three weeks in, in January. We had a backlog of bugs. We are roughly $15 million GMV trying to like 10X this year, and there are so many initiatives that are higher ROI. So what eventually gets dropped are tiny little bugs, the really annoying ones that our developers just never get to. So built autonomous bug fixer in less than four days, literally 25 plus bugs fixed in week one and shipped. And the benchmark pass rate, there's a debate as to whether this is right or not. Basically, one shot fixing of bugs is like sixty to seventy percent right now. And the biggest issue here, or the learnings, is how we can feed it enough context such that it's self-improving. Second business use case. This is my favorite. Talking and teaching about AI agents in plain English. I realize that's a edge 'cause no one has a clue as to what's going on in this whole world. And I realize it's really fun to teach people in really simple English. So I connected the dots. So as Yhangry, we offer private chefs to customers. We also partner with vacation rentals so they can offer their guests private chefs. So there's like a B2B partnership here. I pitched to conference slots saying, instead of pitching Yhangry, I'm going to teach everyone how to build AI agents in 30 minutes. Everyone was like, "Oh my God, done." So I got $50K worth of conference slots for free. Within those conference slots, the deck literally runs through affiliate link integration and demo of the new Yhangry AI product. It's like win, win, win, win, win. And then everyone after the event, because they take screenshots, I just say, "Screenshot the slide, chuck it into Claude code, learn about it later." Everyone ends up posting on LinkedIn saying, "Oh my God, that was great value." So this has been a very fun, random business use case, but growth opportunities are insane. Our product now feels very old school. Finding a chef, booking blah, blah, blah. It's just so many steps and so much back and forth that we are losing a lot of people. And we have so much data. What do chefs respond to? What do customers like? There is no reason for this back and forth to take days. Like, chefs shouldn't have to send the same stuff back and forth to a customer, to millions of customers. We have all the data. We just never knew how to match them instantaneously and kind of really make sure that the one-shot match is really good. That's what Yhangry AI is for on the customer side. And on the chef side, chefs are wasting so much time on admin, like sending the same stuff over and over and over again. So our AI product will just make it very simple for them. We'll just do everything for them. It's like Claude for chefs. So that's kind of like the vision, and that's what we're building for the product side, and we've already validated this on chefs. Like, everyone wants it. It's just not good enough to ship [chuckles] . And it's like the chef side is too divergent. Like, chefs could ask anything. Like some chefs are very by the book, "I want to do X or Y," and some chefs are like lost in the matrix. They're like, "Oh my God, this is so good." And we are just trying to figure out what's like the MVP to launch and how to figure this out. Okay, now the next slides are gonna be quick. Every single area in our company, people are doing AI agents or making sure their workflows are AI native. Just quick rewind. I had a baby three months ago. I got hooked on OpenClau. I was using voice, Telegram, to the point where I nearly separated from my husband 'cause I was like, "I love this." We're okay now. I fired my tech lead 'cause I realized he did not know what skills was, and he was the ceiling in our company. And I rehired our new head of engineering all within a week. In March, we are really all in. This is like the org question Tom was asking, and I'm in the middle of this figuring it out, and I probably need to kill my human empathy a little bit more, a little bit faster. Um, learnings. Weekly agentic labs. Everyone has a different learning curve right now, and it's really kinda wild. Like, people are saying they're building agents, but you just don't know how good they are. So we're trying to do this, we as in me and then the head of engineering together, like literally drawing out stuff and getting Claude code to just like, you know, use all the transcripts and chuck it back in a nice diagram so everyone understands after. From the marketing growth. This is my favorite. It really like-- I finally understand what founder brand means. You just do things you are really good at, get a lot of attention, and funnel it back to your company. That's it. And my domain knowledge is translating my AI Columbia degree from 2013 and ability to grasp concepts in plain English. Like that's it. And I'm giving it away for free, and it's all coming back to Yhangry

Episode duration: 4:55