Superlinear Academy

Cohort 9

Cohort 8

Cohort 1

Cohort 2

Cohort 4

Cohort 5

Self-paced cohort

Cohort 3

Cohort 7

Cohort 10

Cohort 11

Cohort 12

Cohort 13

We show you what's possible with AI, then take you there.

Founder | Meta/Amazon/Statsig/Tencent

Yuzheng Sun, PhD

Prev. Evangelist @ Statsig (Acq. by OpenAI), AI Director @ Tencent; Meta/Amazon

Applied Scientist | yage.ai

Yan Wang, PhD

Applied Scientist @ Samsara | Olympic Torchbearer | Ex Pinterest/Microsoft | Ex Tech blogger with 100k+ subscribers

Build with AI: Mindset, Habits & Tools for the Next 20 Years

Data Scientist

Analytics Engineer

Senior Associate

Data Engineer

Senior Software Engineer

Software Engineer

Analyst

Staff Engineer

Sales Engineer

Statsig

Amazon

Boston Consulting Group (BCG)

Tencent Games

Is this you?<ul><li>Chatting endlessly with an “AI intern” instead of designing the context with clear principles.</li><li>Blaming “hallucinations” when the real issue is poor orchestration.</li><li>Trying to run an entire project from a single, overloaded chat window.</li><li>Treating AI like witchcraft and prompt spells instead of engineering.</li></ul>AI gets better every day. But when it’s time to ship something serious, you don’t yet have a reliable way to make AI show up for you.The problem isn’t AI. It’s how you’re using it. This course exists for that exact gap.AI is a new computing paradigm. To use it well, you have to:<ul><li>Unlearn old patterns</li><li>Install new mindsets and habits</li><li>Build a meta-learning system that can carry your career for the next 20 years</li></ul>That's hard, and we take it seriously. This course is designed to teach you how to fish. On top of lessons and projects, we built two unique tools:<ol><li><a target="_blank" rel="noopener noreferrer nofollow" href="https://youtu.be/dXjcy6XSN_Y">A build platform</a> that consolidates key APIs and lets you deploy your projects with one command—so you can build fast, not fight boilerplate.</li><li><a target="_blank" rel="noopener noreferrer nofollow" href="https://www.superlinear.academy/c/share-your-projects/">A lifetime community</a> where you can share projects, ask questions, and get unstuck long after the course ends.</li></ol>The end of our course is the beginning of your learning, and we will be with you.

Learn AI for the Next 20 Years, Not the Next 20 Days

Engineer Robust Context to Eliminate Hallucinations

Curate "Sterile" Contexts for Peak AI Performance

Architect Multi-Agent Systems to Conquer Complexity

Define Success Criteria to Transmute from Operator to Manager

Build "White-Box" Evaluation Tools to Visualize Quality

Design Self-Healing Loops for Autonomous Optimization

Software Engineers &amp; Developers Ready to build reliable, scalable AI systems integrated into production workflows.

Technical Product Managers &amp; Strategists Seeking the hands-on knowledge to scope, architect, and guide AI initiatives with confidence.

Data Professionals &amp; Automation Experts Who want to transform proof-of-concept scripts into durable applications that multiply productivity.

Production Grade AI Evals by Braintrust.dev

Mengying Li

Mengying Li, Head of Data at Braintrust.dev

Understand why they're essential for shipping reliable AI products

What evals are

Learn to translate “what good looks like” for your AI product into concrete, measurable signals that align your team.

How to define meaningful quality metrics

Design lightweight but robust evaluation suites that expose true product failures instead of vanity score gains.

How to build high-signal evals that catch real regressions

Understand the strengths of each and how to combine them into a continuous feedback loop from development to production.

When to use offline vs. online evals

Set up automated gates and dashboards that let you move fast without breaking quality.

How to integrate evals into your shipping workflow

How I Won First Places in the YC and OpenAI Hackathons?

Riley Shu

YC/OpenAI Hackathon Winner | PM @ Fortnite

To win a world class hackathon, you need good ideas that came from real problems. Pre-hackathon preparation is crucial

Strong idea = useful + unexpected solution

Don't waste time on communication or coding conflicts. Ideal structor is one coder, one generalist, and one support.

Small teams outperform large ones

Avoid slides. Focus on live demo. Design for "wow effect" and clear user experience. Present impact clearly.

Demo-frist development approach

AI tool integration. How to manage expectations and randomness. And more!

More practical tips

The MCP Myth: Do You Really Need a Unified Agentic Protocol?

Staff Applied Scientist @ Samsara

Understand the hidden business and technical motivations behind the push for a unified protocol in the LLM ecosystem.

Why was MCP created in the first place?

Gain practical insights into making informed choices about MCP adoption based on real-world trade-offs.

How can you strategically leverage MCP?

Discover scenarios where adopting MCP could slow you down rather than accelerate your development.

When should you actually avoid MCP?

Level up team productivity with personalized AI agents

Leverage AI as a secretary and coach to help you communicate better, and become a better communicator.

Build AI agents to assist written and verbal communication

AI-enhanced communication empowers teams to adopt innovative strategies to boost collaboration.

Unlock new strategies to overcome  team challenges

Have good judgement on good and bad prompts. Use tools to refine, reuse, and redistribute your prompt.

Master technical skills to build personalized AI that works

Discover practical skills and strategies to improve collaboration.

Join with your team and learn together

Why is your ChatGPT lazy and dumb? (and how to fix it)

Because of a common misconception on how ChatGPT works internally

Understand why ChatGPT could be lazy and dumb

A practical trick that you can apply immediately, for most tasks, everyday

Make all your prompts more effective

Learn why the trick is so effective by understanding Large Language Model research

Learn about the research behind it comprehensively

AI-Assisted Programming for Non-Devs

Build a better-than-commercial photo resizing app in 5 mins (basic python skills required).

Build a prototype with ChatGPT and Python

GUIs used to be the way to access compute, but now GenAI has enabled accessing compute through chat.

Understand a new paradigm of accessing computes

Build often and fast to overcome the mindset bottleneck of productivity.

Learn directly from Superlinear Academy