Oral Cancer Detection AI, Vibe-Coding a Throwaway Micro App, and a Tsunami of Shiny Garbage

Happy Thursday.

In today’s issue:

• Angel Deals of the Week | 134 Deals, 93 Networks

• Bookmarks I think you’ll enjoy

• The best nugget from my conversation with Startup CTO Expert, Igor Belagorudsky

🔥 Angel Deals of the Week

Angel funding rounds announced in recent weeks, compiled from public sources. These deals represent the elite few that survived an angel network’s vetting process. Note: I have not personally analyzed these companies and am sharing for informational purposes only.

📊 Angel Network Investment Tracker: 134 Deals, 93 Networks.

BrightHeart | AI-driven prenatal ultrasound platform

Participating Groups: IDAHO HealthTech Club via SideAngels

BrightHeart is developing an AI platform for prenatal ultrasound that delivers expert-level fetal heart screening with high accuracy. The FDA-cleared software integrates into routine workflows to improve exam completeness, reduce diagnostic errors, and support clinical decision-making. The French company recently closed an $12.8 million Series A co-led by Odyssée Venture and GO Capital, with participation from IDAHO HealthTech Club via Side Angels, Mussallem CHD Alliance, Lift Value, and founding investor Sofinnova Partners. Funds will support U.S. commercialization and European expansion.

Cécile Dupont | $12.8M Series A | Paris, France | January 2026 | Source

OraLiva | AI-powered oral cancer detection

Participating Group: RTP Angel Fund

OraLiva is developing an AI-powered diagnostic platform for early oral cancer detection, combining proprietary microfluidics with models trained on 13 million cellular images. The test claims to achieve 97% accuracy in identifying precancerous changes. The company recently closed a $2.0 million oversubscribed Seed round led by angel Dr. Preetpal Sidhu, with participation from DCVC, RTP Angel Fund, and NYU Innovation Venture Fund. Funds will support commercial launch and CLIA lab opening in Durham.

Spencer Price | $2M Seed | New York, NY | October 2025 | Source

Astrocyte Pharmaceuticals | Neuroprotective therapies for brain injuries

Participating Group: SideCar Angels

Astrocyte Pharmaceuticals is developing neuroprotective therapies for concussions, traumatic brain injuries, stroke, and Alzheimer's disease. The clinical-stage company's lead candidate AST-004 is currently in Phase 2 trials and recently received a $3M NIH grant to develop an oral formulation. AST-004 activates astrocytes - the brain's caretaker cells - to protect neurons and promote recovery. The company recently closed a Series B round with participation from SideCar Angels.

William Korinek | Cambridge, MA | December 2025 | Exclusive

📣 Have an Angel Deal to Announce?

🔖 Bookmarks

1. 🪖 Why “Palantir for X” is Not a Good Pitch: Thought piece from Marc Andrusko at a16z on why most startups copying Palantir's playbook are setting themselves up for failure. I literally saw two deals with this framing in the last 2 weeks. See also what the heck does Palantir actually do.

2. 🤯 The Year of the Mega IPO?! NYT reports SpaceX, OpenAI, and Anthropic are all prepping to go public. Which could mint 16,000+ new millionaires in Silicon Valley.

3. 📱 The Rise of Throwaway Micro Apps: TechCrunch profile on how non-developers are vibe coding “throwaway” personal apps instead of buying them.

🥇 The Nugget: My Top Takeaway from a Conversation with Startup CTO Expert Igor Belagorudsky

AI is Churning out an Ocean of “Shiny Garbage” Startups 🗑️

In the last month, I’ve seen at least a dozen “startups” shamelessly pitching what is functionally a CustomGPT.

As an angel group operator, it’s exhausting. But what I struggle with most is that these founders often seem to genuinely believe what they’ve built is worth scaling to the moon.

Claude Code. Cursor. Lovable. Even CustomGPTs. All these incredible tools raise the waterline of what's technically accessible to non-technical founders.

Which is great until you find yourself surfing a tsunami of awful startups that seem increasingly sophisticated on the surface.

Founders can now ship MVPs in days that would have taken months just two years ago, but evaluating code quality, architecture decisions, and whether something actually solves a real problem hasn't gotten any easier. Non-technical founders can Lego together cool-looking products without understanding if they're built on something legit.

Basically this:

Takeaway: As AI makes it easier to build, double down on the fundamentals. Don't let yourself be dazzled by a pretty demo that took 3 hours to vibe code.

Want more? Check out my full conversation with Igor 👇

Listen now on Apple Podcasts, Spotify, YouTube, and more.

🕵️ Looking for an Angel Group?

Finding the right angel community can be tough.

Answer 12 quick questions and I’ll curate 3 groups you might enjoy.

Even better - if one of them feels like a fit, I’ll personally make the intro.

Until Next Week 👋

Thanks for reading - have a great week.

-Andrew

P.S. If you enjoyed this post, could you do me a quick favor? Hit the "like" button or leave a comment with your thoughts. It may not seem like much, but it really helps me out a ton.

If you’re finding this newsletter valuable, share it with a friend, and consider subscribing if you haven’t already. You can also find me on LinkedIn and X.