Andrew Ng

Who they are

Andrew Ng is the founder of DeepLearning.AI — co-founded Google Brain in 2011, Coursera in 2012, and built Machine Learning Yearning as a freely distributed practitioner's guide.

Person

Andrew Ng launched DeepLearning.AI in June 2017 as a self-funded bootstrapped venture, right after leaving Baidu — no VC, just the platform. His PhD came from UC Berkeley in 2003, his MS from MIT in 1998, and he cut his teeth at Stanford running autonomous helicopter research and the STAIR project, which produced the Robot Operating System (ROS). From Stanford he co-founded Google Brain in 2011 — the deep learning research arm that put large-scale neural networks on Google's map — then served as Chief Scientist at Baidu before stepping back to build his own companies. The companies he's started tell the arc clearly: Google Brain (2011, acquired), Coursera (2012, co-founded with Daphne Koller, now over 100 million learners), Landing AI (2017, industrial AI and Document AI), DeepLearning.AI (2017, AI education), AI Fund (2018, $175 million venture studio co-founding AI startups), and Profitmind (Chairman and Co-Founder). He also backed Kira Learning, an AI education startup aimed at integrating AI agents into K–12 classrooms. The through-line is a single consistent bet: AI's value compounds through distribution and education, not just research. He writes The Batch — a weekly newsletter on AI trends — publishes at andrewng.org, keeps a Medium presence at medium.com/@andrewng, and authored both Machine Learning Yearning (free practitioner's guide) and the AI Transformation Playbook (2018); in 2026 he is organizing AI Dev 26 in San Francisco and joining a keynote panel at Ai4 2026 with Geoffrey Hinton and Fei-Fei Li.

Company

DeepLearning.AI reported $52.8 million in annual recurring revenue in 2025, with a disclosed valuation of $158.4 million — notable because the company has never raised outside venture capital since Ng self-funded it at launch in 2017. In 2026 DeepLearning.AI launched a public GitHub repository with developer tools, course artifacts, and a master course catalog, signaling a move toward deeper developer-community engagement. Earlier, in January 2024, it partnered with CircleCI to launch a free short course on automating large language model evaluations. The company collaborates with AWS, Google, OpenAI, Anthropic, Snowflake, Arm, and Red Hat to build and distribute courses, positioning it as the curriculum layer across the major AI platform providers.

Market

DeepLearning.AI competes in the AI education and technical training space against ai.google, d2l.ai, Pluralsight, Codecademy, and NxtWave, as well as platform-adjacent pressure from OpenAI and Anthropic who are building their own developer education programs. The broader AI industry is navigating sharp regulatory divergence — the EU pushing stricter oversight while the US favors a deregulated, market-driven approach — plus intensifying US–China geopolitical competition that affects partnerships and market access. DeepLearning.AI sits in a structurally interesting position: bootstrapped and profitable at $52.8M ARR while the model providers it trains people on are raising at multi-billion-dollar valuations.

Network

Ng's closest named collaborator is Daphne Koller, with whom he co-founded Coursera. He has a recurring public relationship with Fei-Fei Li — most recently a keynote conversation at Stanford HAI on healthcare and AI, and a joint panel at Ai4 2026 — and Geoffrey Hinton is also on that Ai4 2026 stage with him. DeepLearning.AI's institutional relationships run through AWS, Google, OpenAI, and Anthropic as course-content partners.

Daphne Koller· Co-founder of Coursera
Fei-Fei Li· Co-Director, Stanford HAI; AI researcher
Geoffrey Hinton· AI researcher; Ai4 2026 co-panelist

How they likely show up

Long tenure building DeepLearning.AI since June 2017 as its sole founder with no outside capital → operates with full autonomy and makes long-arc bets without board pressure.
Hybrid role-type pattern — simultaneously running DeepLearning.AI, AI Fund, Landing AI, and chairing Coursera → comfortable holding multiple operating contexts at once; don't expect a single-threaded conversation.
The Batch newsletter (weekly, sustained since launch) → favors consistent, high-cadence communication over big-bang announcements; will have already formed a view on whatever AI topic you raise.
Machine Learning Yearning and AI Transformation Playbook both distributed free → sees broad access and practitioner utility as the primary metric, not monetization of IP.
Hosting and organizing AI Dev conferences (2025 NYC, 2026 SF) rather than just speaking at them → prefers to shape the agenda, not just respond to one.
Coursera co-founded with Daphne Koller at academic scale, Google Brain built inside a giant, AI Fund structured as a studio → repeatedly builds institutional infrastructure around a thesis rather than just funding point solutions.

Conversation tips

→ Reference a specific issue of The Batch or a recent letter at andrewng.org — he writes weekly and will know immediately whether you've read him or are just name-dropping.
→ Come with a concrete view on AI education access or the regulatory divergence between the EU and US — these are live themes in his public writing and he engages with specifics, not platitudes.
→ Ask about the bootstrapped-at-scale model — $52.8M ARR with no VC is a deliberate structural choice, and he'll have strong opinions on why.
→ Don't treat DeepLearning.AI as just a MOOC company; it's the curriculum layer across AWS, Google, OpenAI, and Anthropic simultaneously — that positioning is the interesting thing to probe.
→ If you've used or built with ROS, mention it — the STAIR project at Stanford is where that came from, and it connects his research identity to his practitioner-focused present.

Toolbox

Openers

Open on the Ai4 2026 keynote panel — he's sharing a stage with Geoffrey Hinton and Fei-Fei Li, a rare alignment of the three most publicly prominent figures in the field's history, and it's a natural entry into what he thinks the moment actually means.
Reference the 2026 GitHub repository launch — DeepLearning.AI pushing developer tools and a master course catalog publicly is a meaningful shift from a course platform to a developer-infrastructure play, and he'll have thought hard about why.
Bring up Kira Learning — he's backing an AI-agent-in-K12 startup at the same time he's running a technical professional education company; the two bets together reveal something about where he thinks AI education goes next.

Discovery questions

DeepLearning.AI partners with AWS, Google, OpenAI, and Anthropic simultaneously — how do you keep the curriculum neutral when the platforms you're teaching on are competing fiercely with each other?
You bootstrapped DeepLearning.AI to $52.8M ARR without outside capital — at what point, if any, does that constraint start limiting what you can build, and what would have to be true for that calculus to change?
The AI Dev conferences you're organizing bring together Anthropic, Snowflake, Arm, and others under a DeepLearning.AI banner — is that a community play, a business development play, or something else entirely?

Avoid

Don't pitch him on the idea that AI education is a commodity threatened by model providers' free tutorials — he's thought about this longer than almost anyone and will find the framing reductive.

Make it yours