Vantura

The idea

I used to tutor, and one trend that I began to realize while talking with the parents of my students was that most of them had super busy mornings, no time for anything else. Between getting ready for work, packing lunches, and getting kids out the door, they rarely had time to keep up with the news and their own interests.

That conversation stuck with me, and I kept thinking about what it would look like to have an efficient tool that could update you every morning, like an executive assistant. Something that isn't as distracting as social media, or anxiety-inducing like the news, or ad-filled like the radio. Around the same time, I also began to pay attention to some new apps built on top of the iMessage API, things like Poke and Tomo, and it clicked that this could be the right home for the idea. If we somehow had a personal assistant specifically for news and keeping up to date on information, that'd be really cool. Additionally, with the amount of AI slop out there, I wanted the app to feel like a text from a friend updating you on what's been happening recently.

After some convincing, I convinced my cofounder to pivot into this idea, and as the technical guy, I began to build out the idea.

How it works

The first thing a new user does is go through onboarding (there's a clip of it in the video above), where we learn what they actually want to follow: specific topics, sports teams, stocks, etc.

From there, we go pull the news. We use a mix of News APIs and other sources, along with Grok, which we leaned on to search by topic. It wasn't the most efficient approach, because the News APIs often produced junk. We hadn't built a clean filter yet to separate the useful articles from the noise in the News API results, so in the beta we just let Grok do that work for us, and it held up well enough.

Every morning at a specified time, we deliver custom news to each user through a set of visual news cards they can scroll through and react to right inside the thread. We also deliver a podcast style script that gets sent over to ElevenLabs, which is what actually produces the audio briefing that lands in their inbox each morning. If the user has any follow up questions, they can ask Vantura, and it will reply with its knowledge base, or if prompted, will conduct a web search for the most recent information on what you're looking for.

What we learned

The hardest part of all of this had nothing to do with news or audio. It was getting a language model to behave. Model output is stochastic by nature, so pinning down something consistent and safe enough to send to real people took a lot of trial and error. I spent a long stretch deep in prompt engineering, fine tuning, and RAG before we found something that worked: a loop with guardrails built in, regex checks plus an LLM acting as a judge over its own output (we used DeepSeek for that since it was cheap). Getting Vantura to call the right tools at the right time was its own version of that same fight, since being right *most* of the time isn't good enough when someone is trusting you with their morning.

Along the way we picked up some real signs that we were onto something. Founder University accepted an earlier version of Vantura into their pre seed program, and even though the product had changed a lot by the time we actually shipped, that early validation helped us sharpen our pitch and get comfortable in investor conversations. ElevenLabs also gave us about five thousand dollars in credits, which is really what made the daily podcast possible in the first place. We also received a microgrant from our University to continue developing the product.

What's next?

Since the launch earlier in June, I have since stepped back from Vantura to pursue personal research goals. I had a blast building the product, and it was a great learning experience to learn about agent harnesses, Langchain, and using language models. I'm excited to see wehre Vantura heads in the future, and I hope it continues to find its footing. I also learned a lot about starting, running, and maintaining a startup. Building a real product and talking to users is no joke, let alone pitching yourself to investors!

Onto the next.