What I Learned From Living Alone
Takeaways from building in SF, how to go viral, and more.
My goal with this newsletter is to give you the top insights from my life.
If you’re a new reader: I’m a teenager interested in unlocking human potential through better education. I’m building conversational teddy bears for early childhood education, and previously led AI engineering for a venture studio.
This issue includes:
What I learned from building in San Francisco
Early demo of the talking teddy bear
Instructing LLMs to scrape data
What I Learned From Living Alone and Building in San Francisco
I spent most of my summer at Founders, Inc.'s residency for university students and new grads building their own companies.
Above is the group picture from our demo day at the end!
If you're planning on going to San Francisco for a while, here's my advice:
1/ Be open-minded, but don't get distracted from your mission
This city has thousands of events a year (sometimes 10 a day). It's easy to get distracted going to event after event and then justifying your time spent by calling it "networking."
I personally think "networking" is overrated.
I think focusing on doing great work, and tweeting about it so the public sees + sending high-signal and targeted cold messages will get you further than sacrificing your ability to do great work by constantly going to events.
2/ Remember that SF and its communities are bubbles
I feel like a lot of founders in SF have trouble socializing with non-founders from the rest of California. Because your environment shapes who you are. When you're in a community with such like-minded people, you lose out on insights from the rest of the world.
A good analogy here is reading. A lot of "hustlers" will only read self-help books and stay clear of fiction. If you're like that, let me offer you this quote:
“Reading isn’t about putting information in your mind. Reading is about sparking a fire in your mind” - Naval Ravikant.
Books aren't about getting info, it's about sparking new thoughts in your mind. I think entrepreneurship is the same.
3/ Pay it forward and be kind
I can't prove that karma is real, but what I can say is this: SF is small. Both in population and in size (it's only 7 miles x 7 miles). Word gets around quick. Before even thinking to wrong someone else, think about how what you do will come back to you. Your reputation isn't your number of followers - it's what people say about you in the group chats and the other places you can't see.
Early Demo of The Talking Teddy Bear
I released an early demo of the talking teddy bear I’m building. The conversations I’ve been having with journalists, investors, operators, parents, and kids has been super energizing and the feedback has been great.
I’m spending most of my time learning how to go viral and taking shots on net to grow our waitlist. When I’m not testing new short-form content concepts for vitality, I’m testing the teddy bear with families or improving the product’s UX. Shoutout to
for being an amazing mentor through this.
Some insights on going viral:
Retention is everything. Your first 3 seconds should convince viewers to stay to the end.
Use action words in your hook. Too many people start their videos with something like “This is X.” So what? Give people a reason not to scroll by creating a story by using action words.
Example: “Pranking my family with X.”
Look at what other brands are doing. They’re mostly all just copying each other, successfully.
Instructing LLMs to Scrape Data
I’ve had the dream part-time job this summer helping Simple Ventures’ portfolio founders improve their MVPs through software.
One of the startups I worked with is Canadian Longevity. It’s a service that gives you a simple blood test to discover 60+ of your biomarkers, get personalized coaching based on your results, and calculate you an overall longevity score.
The Canadian Longevity team has been manually copying data from the lab’s patient result PDFs into their spreadsheet so its underlying algorithm could calculate the longevity score. I automated that process with an experiment instructing LLMs to scrape data from a PDF, format it in JSON, and then write that data to the spreadsheet in the right places.
The experiment was a success.
I’ve open sourced the GitHub repo so you can understand how it works/embed it into your own software. You can fork the repo and try it for yourself!
What’s next?
Launch orders for the teddy bear this month
Wrap up work at Simple Ventures
Thanks for tuning in!
Insightful Krishiv
Let’s goo Krishiv!!! So glad to see you putting yourself out there in the direction of your vision!