The Horowitz Andreessen Academy

The Horowitz Andreessen Academy (HAA) is a private, full-time, in-person school in San Francisco for young builders coming out of high school. a16z announced it on September 22, 2026, incubated it and is investing $35 million to launch it. The Academy is a separate, for-profit company led by founder and CEO Gagan Biyani (cofounder of Udemy and Maven); Marc Andreessen and Erik Torenberg sit on its board.

The first class, the Founding Class Fellowship, starts in September 2027 with about 50 students. The year runs September to April on the San Francisco campus, then a project abroad in May (with a $5,000 travel budget) and a co-op from June to August. Every student spends at least three months working full time inside a company. Students split their time between co-ops, practitioner-taught courses (examples include AI Inference Engineering, The Craft of Selling and Venture Capital) and self-directed Pursuits such as a startup, research or art project.

Founding partners Anduril, Anthropic, Coinbase, Google, Meta, NVIDIA, OpenAI, Palantir, Replit and Stripe help shape the curriculum and offer co-ops, and more than 40 other companies are hiring partners. Guest lecturers named by The Academy include Sam Altman, Jensen Huang, Brian Armstrong, Garry Tan and Travis Kalanick. There is no tuition; The Academy is not accredited and does not award a degree.

Designed for Students. The program is in person. Participation is generally full time.

• High school diploma or equivalent completed by August 2027.
• Not eligible if you will have completed more than two years of full-time college after high school by August 2027. Current college students are otherwise welcome.
• Priority goes to graduating high school seniors.
• International applicants are welcome; US citizenship or residency is not required, but visas are not guaranteed.
• No age requirement. Students under 18 at the start need a parent or guardian to co-sign enrollment documents.
• No coding requirement: builders of software, products, research, communities, audiences or events all qualify.

The Academy looks for agency: people already in motion who act on what captures their attention. Admission is heavily weighted on proof of work rather than grades.

Stage 1: share your work. Submit contact information, a link to a personal portfolio website, an essay, a short video, test scores and your high school transcript (transcripts and scores are for records). No recommendation letters at this stage. AI may be used to move faster, but not to write your story.

Stage 2: continue the conversation. Selected applicants are invited on a rolling basis to further conversations about their work.

Early decisions arrive by late January 2027. Applying early does not commit you to attend. The class is expected to be about 50 students.

• No tuition for the Founding Class Fellowship.
• Courses taught by founders, investors and practitioners, plus guest lectures and mentor office hours.
• At least three months working full time inside a company, with co-op options at founding and hiring partners.
• A $5,000 travel budget for a three- to four-week project abroad.
• Access to compute and hardware through founding partners.
• A peer group of about 50 builders in San Francisco.
• Need-based scholarships for some living expenses may be available.
• Students keep 100% of their IP; no equity, income share or right of first refusal.

• September 22, 2026: The Academy announced by a16z.
• November 5, 2026: early application deadline.
• Rolling: invitations to stage 2.
• By late January 2027: early admissions decisions.
• Early 2027: regular admissions round, only if spots remain.
• September 2027: Founding Class Fellowship begins.
• Program year: SF campus September to April, abroad in May, co-op June to August.

The Academy is new; the first class starts in September 2027, so there are no outcomes yet. Early applicants featured on the site include builders who raised $100K for an enzyme engineering startup, built an AI book indexer with six-figure ARR and co-created MIT's first class on AI agents.