Hello builders,
This week, the gloves came off.
China just dropped Kimi K2, a trillion-parameter LLM. But, the icing on the cake…
They made it open source.
It’s one of those holy sh*t moments, right?
In this week’s newsletter.
New Startup Spotlight: Health OS is targeting the $4T US healthcare market.
AI News Recap: China’s leading the race (still), Replit playing cloud hopscotch, and AI malware getting smarter.
Recent Job Postings: A founding engineer role with up to $1M in equity on the table.
Let’s go.
18
Engineer spots left this month
Each month, I work with a curated group of cracked 0→1 engineers — the kind of founders founders want to hire or co-found with.
As of this issue, only 18 spots remain.
🔍 Startup Spotlight
HealthOS
HealthOS wants to flip healthcare from reactive crisis management to proactive, personalized care.
They’re building a health operating system that uses DNA, biomarkers, and real-time data to predict health issues before they happen.
Instead of treating patients like population averages, they create a digital twin of your health that evolves with every test and habit change.
It’s an ambitious bet on the idea that prevention, not treatment, is the future of medicine.
There’s huge upside for engineers who want to work on healthcare that’s both deeply technical and genuinely personal.
TL:DR
- Building a personal health OS targeting the $4T US healthcare market
- Uses genomics + real-time data for predictive health insights
- Blends consumer UX with backend data engineering
- Founding engineer role → high ownership and equity upside
- Backed by Sunflower Capital (Cohere, Retool) and BoxGroup (Ramp, Oscar Health)
Megan’s Take
Money, talent, and tech are all flowing into health right now. HealthOS is betting on personalized prevention. It’s risky, but if they’re right, they’ll carve out a chunk of the $4 trillion healthcare market. This is worth a look.
🤖 AI / Industry Recap
🔗 Moonshot AI Open‑Sources Kimi K2
China’s Moonshot AI released Kimi K2, a mixture-of-experts model with 1 trillion parameters (32 B activated) capable of complex reasoning. It claims to outperform open-source rivals like DeepSeek V3.
🛠️ Replit Joins Microsoft Azure
Replit announced a non-exclusive partnership with Microsoft Azure Marketplace while growing ARR from $10 M to $100 M in just six months.
🏛️ AI malware can now evade Microsoft Defender
Security researchers at Outflank trained an open-source LLM (Qwen 2.5) to generate malware that evades Microsoft Defender about 8% of the time.
Megan’s Take
Big tech’s finally sharing some of its toys, and it’s giving regular engineers a shot at building stuff that used to be out of reach. It’s exciting, but also a bit wild.
Recent job posting in
NewYork
jobTitle: Staff Software Engineer_
}
// TLDR: This is heavy backend and infra work. You’ll be designing ledgers, reconciling billions in transactions. Think Stripe for logistics, with big equity upside.
// Company: AtoB;
// Location: NewYork, (Hybrid - Mon-Thu at various office locations (SF, NY, LA);
// Base: $225k - $300K; (Equity = $600K – $1M)
// Funding: $182M raised from General Catalyst, Elad Gil, Mastercard, and others
}
OUTPUT:
Recent job posting in
NewYork
jobTitle: Founding Product Engineer_
}
// TLDR: You’ll be building a personalized health OS that uses genomics and real-time data to catch issues before they become crises. You’d be their first engineering hire after the founders. Big upside with equity up for grabs.
// Company: HealthOS;
// Location: NewYork (On-site)
// Base: $200k - $300K; (Equity = 1-1.5%)
// Funding: Backed by Sunflower Capital (Cohere, Retool) and BoxGroup (Ramp, Oscar Health);
}
OUTPUT:
Recent job posting in
SanFrancisco
jobTitle: Principal Software Engineer_
}
// TLDR: BRM builds tools to help businesses manage contracts, vendors, and spending. This role leads architecture and ships new features in a small team environment. Perfect for engineers who want end-to-end ownership and visible impact.
// Company: BRM;
// Location: SanFrancisco (On-site);
// Base: $200k - $250K;
// Funding: $21.6M raised
}
OUTPUT:
Recent job posting in
SanFrancisco
jobTitle: Applied AI Engineer_
}
// TLDR: Verita AI builds tools for generating training data in tricky fields like design and finance. As their first US engineer, you’ll design internal systems, deploy LLMs, and integrate with enterprise data.
// Company: Verita AI;
// Location: SanFrancisco (Onsite - 6 days/week);
// Base: $200k - $240K;
// Funding: $1M raised
}
OUTPUT:
AMA
Question:
Meg, I keep seeing ‘founding engineer’ roles, but I’m not sure I’m cut out for it. How do I know if I’m ready?
Answer:
Such a good question because “founding engineer” is a sexy title, but a very real lifestyle shift.
Ask yourself this:
Does ambiguity energize or drain you?
Can you move fast and ship even when you don’t have perfect data?
Would you rather build the system or just operate it once it’s stable?
Also, are you ready for an emotional rollercoaster?
You don’t have to be the world’s top coder to be a great founding engineer. But you do have to be able to solve weird, unexpected problems.
If you’re asking this question, I’d say you’re closer to ready than you think.
— This question was from Neil.
Send me your question by replying to this email.
💡
Thought Prompt of the Week
If open-source models like Kimi K2 become readily available, what’s one problem you’d solve that’s “too niche or too risky” for big tech to touch?
Where would you use AI to shake things up, even if it ruffles a few feathers or breaks some rules?
Who would finally get a tool they deserve, because you were bold enough to build it?
Remember, the edge is where the interesting stuff happens.
★ ★ ★ ★ ★
Working with Megan didn’t feel like working with a recruiter at all. Meg didn’t just spam send me roles. We talked about what I was interested in next and was patient with me.