Applications open
Most students never get access to a practitioner who will
sit with their work and tell them what is actually wrong with it.
Real assignments. Real mentor feedback. A published portfolio and a completion certificate for ambitious students at state schools, HBCUs, and regional universities.
Project-based, not lecture-based. Every week builds one business muscle and produces one artifact that goes on your public portfolio.
Follow the Money
Pick a real company. Map how it makes money, who pays, what they pay for, and where the profit actually is.
Who Is the Customer, Really?
Talk to real users. Write a customer research piece. Learn the difference between what people say they want and what they do.
Who's Winning and Why
Map a competitive landscape. Three companies, one market. Explain who is winning and why, not just what they do.
Build Something Real
Create a tangible artifact: a working script, a dataset, a product concept. Something that lives on GitHub or your portfolio.
The Memo
Write a one-page business recommendation. Problem. Context. Options. Recommendation. Risk. No fluff.
Final Presentation
Present a complete case study live. Business model, customer, market, your build, your recommendation. Published publicly.
Not a certificate from a platform. A body of public work and a practitioner who knows your name.
01
A public portfolio
Five published pieces of original work on a personal website you own. Not private assignments nobody reads.
02
Real work on GitHub
Code, datasets, or structured research that demonstrates how you think, not just what you studied.
03
Industry mentor feedback
Weekly 60-minute calls with a practitioner who reviews your work the way a manager would, not a professor.
04
AI Fellows Certificate
A signed completion certificate from the Program Director. Something you can show and point to.
Project briefs for fellows, each inspired by a course at The Horowitz Andreessen Academy or on Maven. Pick one and build it in six weeks. Credit goes to the courses; the briefs are ours.
Build a small AI feature, such as a study-guide generator for one of your classes. Write 30 test cases, including clean examples that should pass, and show one real bug your evals caught.
Inspired by Building and Evaluating AI Products (Hamel Husain and Bryan Bischof), The Academy →Inspired by AI Evals For Engineers & PMs (Hamel Husain and Shreya Shankar), Maven →Take a simple research agent and add the operating layer: tool limits, retries, a human approval step and a log of every action. Measure how often it fails before and after.
Inspired by Harness Engineering (Aishwarya Naresh Reganti and Kiriti Badam), The Academy →Inspired by Building Agentic AI Applications with a Problem-First Approach (Aishwarya Naresh Reganti and Kiriti Badam), Maven →Build an assistant that answers questions about your college's catalog or student handbook and cites the page it used. Check 50 answers by hand and report how often the citation is right.
Inspired by Inside the AI Machine (Hamza Farooq), The Academy →Pick one task and run it on three models. Measure quality, speed and cost per answer, then recommend one and defend the trade-off in a one-page memo.
Inspired by AI Inference Engineering (Abi Aryan), The Academy →Find a task people in your town dread, like filing a permit or a claim. Prototype a product that could not have existed before AI, and write why now.
Inspired by Products That Couldn't Exist Before AI (Anish Acharya), The Academy →Rebuild one everyday app task as a voice or chat experience. Test it with five people and compare time-to-done against the original.
Inspired by The Next Consumer Interface (Olivia Moore), The Academy →Design a two-sided campus marketplace, such as students and tutors. Plan how it gets its first 100 users, then run the first step with 20 real people.
Inspired by How Networks Take Off (Andrew Chen), The Academy →Interview ten people who run a local business. Find one painful, repeated job and write the one-page memo: who has it, what they do today, and what you would build.
Inspired by Finding Ideas Worth Pursuing (Mar Hershenson), The Academy →Take an AI product you use. Map it layer by layer, from data and context to the interface, and name what would be hardest for a competitor to copy.
Inspired by AI Systems Thinking for Product Managers (AI Fellows program director), Maven →Essays on agentic systems, AI infrastructure, and how practitioners actually build with AI.
Fifteen years working across engineering, product, data, and AI at enterprise SaaS companies, early-stage startups, and consulting engagements. I know what companies look for when they hire, what business decisions look like from the inside, and what separates people who advance from people who stall.
Most of that knowledge is not taught in school. It lives in conversations between practitioners, the kind that happen in certain rooms and don't happen in others. AI Fellows is an attempt to change that.
Every week I review your work the way a manager would. I tell you what is sharp, what needs sharpening, and what a real business audience would think. That feedback compounds.
Cohort 1 starts July 7, 2026. Limited spots. Remote. Free. The only requirement is commitment: show up every week, do the work, publish it.
Send an applicationEmail angshuman@ai-fellows.org with your name, college, major, year, and one paragraph on why you want this. No resume required.