loading...
Dexa surfaces expert opinions from vetted podcasters – the kind of trusted, personalized perspective no general-purpose LLM can replicate.
ENTRY ANGLES
Replace AI-synthesized opinions with direct citations of specific individual expert views · Layer live human experts on top of AI-powered research/answer finding · Position specific human recommendation as more trustworthy than general AI synthesis
VERTICALS
CAPABILITIES
Access to specific domain experts or practitioners willing to provide direct recommendations, AI tools for finding/synthesizing source material efficiently, Ability to clearly communicate the trust/authenticity value proposition versus general AI
DEXA FOUNDER
“Google is great for finding facts, but less useful for finding advice and recommendations,”
Dexa describes itself as the place "where curiosity meets credibility." The site surfaces expert opinions on health and fitness, AI and technology, business and finance, and personal development.
The source material is podcasts. Dexa has curated and continues adding episodes from "hundreds" of vetted experts.
On any expert's page, visitors can browse recent podcast episodes – watching or listening in full, or reading AI-generated summaries. The page also surfaces the most frequently asked questions the expert has addressed across their episodes, with AI-generated answer text.
But the most powerful feature isn't browsing – it's asking. Type a question into the search bar and Dexa finds the best answer from its expert database. You can ask the open field and let the AI find the most relevant expert, or target a specific person and search only within their content.
Answers come as short clips from longer recordings – the precise moment where the expert says something that answers your question.
Dexa launched in February 2023. Monthly visitors are around 50,000 – modest numbers, but the startup just raised a $6M first round.
The origin story is instructive. The founder is a devoted listener of a particular podcast. In one episode, the host described how magnesium improves sleep quality and gave specific instructions on how to take it.
The founder bought the magnesium. Then realized he couldn't remember the details. Hunting through multi-hour episodes to find one specific recommendation took hours. Out of that frustration, he built a simple AI tool to search within his favorite podcaster's back catalog – and posted about it online. The podcaster shared it. The founder's inbox exploded with requests to do the same for other creators' content. Dexa was born.
The founder's claim is that he's built a new kind of search engine.
"Google is great for finding facts, but less useful for finding advice and recommendations," he explains. "Because in that case, people don't want a faceless text – they want a credible opinion from someone they already trust or could trust. And those opinions are buried inside podcasts and videos that Google doesn't even index."
As one investor put it: "Dexa's ambition goes far beyond replacing Google's ten blue links with two short paragraphs [like ChatGPT or Perplexity]. They're creating a new indexing principle based on individual human perspectives rather than web page content. People search for experience and wisdom, not content summaries."
This taps into something broader: people are fundamentally interested in other people – their personal opinions, their specific experiences. Social networks, blogs, and podcasts all became massive because of this. Dry facts have authoritative, impersonal sources – news organizations, databases. But interpretation, advice, and recommendations flow through trusted individuals.
For a long time, people defaulted to "if it was on TV, it must be true," then "if Google returned it first, it must be authoritative," and now increasingly "if ChatGPT said it, it must be right" But a creator whose work someone follows daily probably carries more credibility than any of those sources for the questions that actually matter in that person's life.
The core insight – human wisdom over content synthesis – is sound.
Quench (covered in summer 2023) spotted the same pattern earlier, starting as a "learning tool for busy professionals" and delivering short expert video clips as answers to user questions. It has since pivoted toward a more general "turn any content into an AI assistant" framing, but the underlying technology is the same.
There's no technological novelty in Dexa's approach – but the most important startup decisions are almost never about unique technology. They're about what you use existing technology for.
Dexa's specific choice: replace ChatGPT-style opinion synthesis with direct citation of specific individual views. That's a meaningful product distinction even if the underlying models are identical.
Interestingly, the same principle can surface in unexpected domains.
Take pet care. The internet is full of general answers to pet health questions. But Dr.Tail, [covered last summer](/review/kogda-mnogo-raznogo-nachinajut-platit-za-drugoe), charges $9.99/month for live consultants to answer those same questions. The consultants themselves use AI tools to find the answers – but people pay because a specific human giving a confident, direct recommendation feels more trustworthy than a general search result.
The principle applies broadly. Dexa uses it for health and business podcast expertise. Quench uses it for professional education. Dr.Tail uses it for veterinary advice.
What product could you build on the same logic? In what areas of your own life do you want the specific opinion of a real person rather than an AI-synthesized general answer? Where does the source material already exist? And how would you convey why that's more valuable than what ChatGPT can produce?