Joe Magyer @Magyer

Watching my wife’s home country play in the World Cup in my home town. Go Uzbekistan!!!

@ntippmann Apple Podcasts: podcasts.apple.com/us/podcast/fro…

Excited to share my conversation with @ntippmann! Nick is a vertical AI and GTM expert so it was great to talk shop with him about AI, how B2B SaaS evolved, and how distribution has changed. Here are some takeaways and insights from the conversation: Vertical AI changes the value proposition from efficiency to execution. Nick’s core framing is that SaaS helped people save time, while vertical AI can actually do the work. That shift changes the buyer, the budget, the pricing model, and the ceiling for how large these companies can become. The real TAM unlock is services and labor, not software. Traditional vertical SaaS was constrained by software budgets. Vertical AI can compete with outsourced services, internal labor, and professional work that previously sat outside the software spend category. That is why Nick sees the opportunity as dramatically larger than the prior wave of vertical SaaS. The best vertical AI companies will not be thin wrappers. Nick’s bet is on companies that combine workflow software, domain-specific context, proprietary data, and agentic execution. The model is not “AI replaces SaaS,” but “SaaS evolves into a system of work.” Incumbents are vulnerable when their strength becomes rigidity. Legacy systems of record often have distribution, data, and customer relationships, but they can also be slowed by old architecture, existing pricing models, and organizational inertia. Nick is most interested in markets where the incumbent is entrenched but not especially nimble. Distribution and trust may matter more in AI, not less. Because software is easier to build, it is harder to stand out. Nick argues that content, community, events, brand, empathy, and domain expertise remain critical. In vertical AI, credibility with the buyer can be a moat before the product moat fully compounds. Outcome-based pricing is becoming the new frontier. AI products are pushing companies away from seat-based SaaS pricing and toward pricing tied to completed work: per resolved ticket, per document, per workflow, per enriched record, or against avoided professional services spend. The best pricing models will align with the value customers actually experience. Forward-deployed services are not a weakness; they may be evidence of where value accrues. Nick sees OpenAI and Anthropic building services-heavy motions as a sign that intelligence alone does not solve enterprise problems. Context, implementation, workflow integration, and judgment still matter — which supports the vertical AI thesis. Seed investors need to look past the demo. In a world where a founder can vibe-code a polished product quickly, Nick focuses on deeper questions: Does the company get better with usage? Is there proprietary data exhaust? Are there reinforcement loops? Can the wedge expand into adjacent workflows? Is there a real vision beyond the initial tool? The wedge still matters, but the expansion timeline is compressing. Nick acknowledges that AI lets companies build more faster, but he still wants a sharp entry point. The difference is that founders may need to move from wedge to broader workflow coverage much earlier than in the SaaS era. GC AI is a case study in founder-market fit plus distribution. Nick backed the company because Cecilia Ziniti brought deep credibility in legal, had early community instincts, and chose a focused wedge in in-house legal. The insight was not just that legal work fits LLMs; it was that in-house legal has distinct workflows, budgets, and pain points from law firms. VCs and founders are often talking past each other. Nick thinks VCs underappreciate how brutally hard execution is, while founders often misunderstand fund math and why “good growth” may still not be enough for the next round. The AI era has raised expectations, especially as investors compare every company to the fastest-growing AI breakouts. Nick’s contrarian view: foundation labs moving up the stack does not kill vertical AI. His argument is that the labs’ move into applications and services actually proves the opposite: raw model intelligence is not enough. Durable value may accrue to companies that own the workflow, the customer relationship, the domain context, and the system of work. Really enjoyed this episode of @InvestNStartups. Thanks to Nick for coming on!

micro1 x @cartesia: The Human Speech Data Project Every language, accent, and speaking style has its own rhythm, texture, and feeling. Building great voice AI means learning from the people who know those details best. That is why micro1 and Cartesia are launching The Human Speech Data Project, an initiative focused on improving the next generation of voice AI through high-quality speech data and human evaluation. The project brings together multilingual speakers, linguists, transcriptionists, native speakers, and voice actors to help: - Capture and evaluate natural conversational speech - Transcribe and annotate audio with a sharp ear for accuracy - Review AI-generated speech for naturalness, correctness, and edge cases - Bring expert judgment into the model improvement process Visit the link in the comments if you’d like to learn more about how you can contribute.

If the US makes the final, get Lee Corso to unretire and have him put on a bald eagle head

“Always go to the funeral”

Jackson Dahl @jacksondahl

2 weeks ago

One of the best bits of advice I've ever gotten, that has consistently held true for the decade since I first heard it from @Mazzeo: When you have bad or difficult news: call the person. Don't text, don't email. Call. Or better yet, meet them in person.

16 164 3K 1.6M 983

Single until Series B makes sense to me now. Conor (@contextconor) finally breaks his silence on the movement, and his reasoning is more nuanced than the X discourse suggests. At the end of our podcast I asked him to expand on his thoughts on dating as a founder. 𝐓𝐈𝐌𝐄𝐒𝐓𝐀𝐌𝐏𝐒 (00:00) Trend Backlash (00:35) Origin Story Explained (01:19) Why Series B Matters (02:11) Not Anti Relationship (02:46) Risks of Dating Mid Build (04:33) Guilt and Founder Singles (05:55) Meta Strategy to Live Best Life (07:18) High Bar Romance (07:51) Wrap Up

Julia Fedorin @juliafedorin

2 weeks ago

A YC fund built a machine learning model to predict the perfect founder. Right school, right company, right age. Conor (@contextconor) met zero of the criteria. That same fund just invested in his round. Nothing about him pattern-matches. Farm kid from a town of 2,000. Dropped

11 10 79 46K 42

Terrible loss and very sorry to hear. Josh gave so much to the Austin community and startup ecosystem.

Austin Statesman @statesman

2 weeks ago

Capital Factory founder Joshua Baer, a visionary force in the Texas technology and start-up ecosystem, died Tuesday night in a private plane crash in Laredo, the organization confirmed to the American-Statesman. "Joshua was a fearless leader, a brilliant partner, and a dear

96 92 690 384K 85

🧵Veria AI autonomously found and demonstrated a critical vulnerability in Aleo, finding a proof forgery in Aleo's snarkVM. The project awarded us the maximum bounty of $65,000 for the find.

Apollo selects Austin as site of second headquarters ft.trib.al/cIlIQRa

Fable 5 (@AnthropicAI) scores 22% and tops the Hedge-Bench leaderboard. Running Fable was roughly 2X more expensive than Opus 4.8 per trial. For an industry where accuracy is mission critical, human judgement isn't going away

Faster decisions start with cleaner data, not bigger teams. Our self-service is live. Any document. Any format. Guaranteed accuracy. No implementation wait. bit.ly/43FuNZc

@RobynAndFawn this summer, it's time to lock in

"Venture-capital investment in Austin reached a record $7.4 bn last year ...The city is now America’s 5th-most active for VC investment, up from 10th a decade ago" economist.com/business/2026/…

For the Analysts: between the sub-15% pass rate and 80% hallucination rate, your job is very safe

Trata (YC W25) @trytrata

3 weeks ago

Today, Trata is excited to present Hedge-Bench, the world’s first benchmark focused on evaluating open-ended reasoning in the finance domain. We curated 102 tasks derived explicitly from the reasoning traces of professional hedge fund analysts working with relevant information

20 29 144 30K 61

@trytrata Love this on multiple levels. Also great to see that humans are still 🔥 at open-ended reasoning in finance relative to LLMs.

Today, Trata is excited to present Hedge-Bench, the world’s first benchmark focused on evaluating open-ended reasoning in the finance domain. We curated 102 tasks derived explicitly from the reasoning traces of professional hedge fund analysts working with relevant information sources. Most benchmarks are geared towards evaluating naturally deterministic tasks (e.g. updating spreadsheets, calculating formulas). The result is a blind spot on the frontier labs’ part to the competency that matters most in the finance industry: reasoning. No frontier model scores above 16% on Hedge-Bench. We observe Claude-Opus-4.8 actually regresses relative to the last generation of Claude-Sonnet. We see hallucination rates go as high as 82% for multi-step reasoning. These implications extend beyond just the finance domain and into other domains where trust in agent reliability is equally critical (e.g. law).

we’re partnering with 50 companies over the next two weeks, each with 50–200 employees, to help improve AI models using real-world company workflows. we believe the companies that helped build modern business workflows should participate in — and very much benefit from — the value created by the next generation of AI systems. for many companies, these partnerships can create a meaningful new revenue stream, often ranging from $100K to $2M+, with opportunities to become recurring over time. our goal is to do this in a way that is privacy-first, low-lift, and aligned with the work companies are already doing every day. if you’re interested in contributing to AI advancement while improving your own workflows alongside micro1 and our frontier AI lab partners, we’d love to hear from you. please reach out to to [email protected] if you’re an executive at a 50+ employee company, excited to potentially partner up!