Welcome, Startup enthusiasts!

Good Morning. It’s July 23, 2026.

🚀 The Startup Khan Newsletter

July 23, 2026 — The AI Bottleneck Moves to Silicon

🧠 Top Story: Etched Raises $300M at a $10.3B Valuation

AI-chip startup Etched has closed a $300 million Series C led by Sequoia Capital, valuing the company at $10.3 billion.

Andreessen Horowitz, SK Hynix, Jane Street, and Diffusion Capital also participated. The valuation has more than doubled since Etched raised $500 million at a $5 billion valuation roughly seven months ago. (TechCrunch)

Founded in 2022 by Harvard dropouts Gavin Uberti, Robert Wachen, and Chris Zhu, Etched is developing full computing systems designed specifically for AI inference—the stage where trained models generate answers for users.

The startup says its first chips have been manufactured, customer testing is underway, and it has already booked $1 billion in orders. Its architecture separates prompt processing from output generation, using specialized chips and shared memory technology intended to improve speed and reduce cost. (TechCrunch)

Why founders should care

Model quality still matters. But as AI usage scales, the economic battle increasingly moves to inference.

Every prompt creates a cost. Faster, cooler, and more efficient infrastructure can turn an impressive AI product into a sustainable business.

👉 Founder takeaway: Your AI margin may depend more on infrastructure architecture than subscription pricing.

💰 Other Startup Moves

1️⃣ Imagi raises $4.5M for AI education

Edtech startup Imagi raised a $4.5 million seed round from investors including Brighteye Ventures, Day One Capital, and Will.i.am.

The company provides K–12 schools with coding and AI-literacy tools. Imagi says it has reached more than 700,000 students across 140 countries and plans to use the funding to expand school-district sales, hiring, and product development. (TechCrunch)

The opportunity is not merely giving students access to AI. It is creating safe, structured environments where they learn to build with it.

2️⃣ Runway becomes an AI-media router

Runway launched Media Router, a system that automatically chooses among image, video, and audio models based on a developer’s priorities—including quality, speed, and cost.

The product sits inside Runway Dev, its API platform used by companies such as Adobe, Cloudflare, ElevenLabs, Expedia, Shutterstock, and Quora. (TechCrunch)

This is a smart strategic pivot: when no single model remains the permanent winner, own the layer that selects between them.

3️⃣ Dassault buys ArisGlobal for up to $2B

Dassault Systèmes agreed to acquire U.S. life-sciences software company ArisGlobal for approximately $1.8 billion in cash, plus as much as $200 million tied to future AI-related revenue.

ArisGlobal’s software supports areas including clinical development, regulatory processes, and patient safety. The transaction is expected to close during the second half of 2026, subject to approvals. (Reuters)

For founders, this is another signal that vertical AI with deep workflow integration can attract strategic buyers.

🧰 Tool of the Day: Inngest

Inngest helps developers run reliable background jobs and multi-step workflows without independently managing queues, retries, scheduling, or workflow state.

It supports TypeScript, Python, and Go, with built-in concurrency controls, throttling, observability, and automatic retries. (Inngest)

Useful founder use cases:

Running AI-agent workflows
Processing webhooks
Scheduling onboarding sequences
Retrying failed API calls
Coordinating long-running jobs

👉 Best use: Add it before mission-critical background processes become a collection of fragile cron jobs.

📊 Trend Check: Infrastructure Valuations Are Outrunning Revenue

Etched reached a $10.3 billion valuation while reporting $1 billion in booked orders—not recognized revenue.

That does not prove the valuation is justified. It does show what investors are pricing: the possibility that inference becomes a massive infrastructure market with room for alternatives to incumbent chip platforms. (TechCrunch)

The founder lesson is broader:

Markets reward credible control over scarce bottlenecks.

🌟 Founder Spotlight: The Etched Team

Gavin Uberti, Robert Wachen, and Chris Zhu pursued dedicated AI silicon when many observers considered the idea too narrow.

Their bet was simple: general-purpose hardware would eventually struggle to deliver the economics required by rapidly growing inference workloads. (TechCrunch)

The playbook:

👉 Identify where industry demand is heading—not where consensus currently sits.

⚡ Quick Win: Calculate Your Cost per Successful Outcome

Do not track only the cost of each AI request.

Measure:

Total model and infrastructure spend ÷ successful customer outcomes

A cheap request that fails is expensive. A costly request that completes a high-value workflow may be excellent economics.

Optimize for outcomes, not tokens.

🧩 Final Thought

Today’s startup winners are attacking different layers of the same shift:

Etched is rebuilding compute. Runway is orchestrating models. Imagi is developing AI-native skills. ArisGlobal is embedding intelligence into critical workflows.

The opportunity is not simply to use AI.

It is to own the bottleneck that everyone else must pass through.

— The Startup Khan

Your 5-minute startup edge — funding news, tools, and founder playbooks.