Welcome to the official DeepLearning.AI YouTube channel! Here you can find the videos from our Coursera programs on machine learning as well as recorded events. DeepLearning.AI was founded in 2017 by machine learning and education pioneer Andrew Ng to fill a need for world-class AI education. DeepLearning.AI has created high-quality AI programs on Coursera that have gained an extensive global following. By providing a platform for education and fostering a tight-knit community, DeepLearning.AI has become the pathway for anyone looking to build an AI career.
Watched DeepLearningAI? Tell other learners what you think below, or read the discussion around the channel.
2:114K views · 2d ago · 2:11
Tabular foundation models use in-context learning to make predictions on new datasets without training a model from scratch. In this short course, built in partnership with Prior Labs and taught by Frank Hutter, Eliott Kalfon, and Tuana Çelik, you'll
1:287K views · 16d ago · 1:28
An agent that calls tools and decides its next step can complete a task. Making that agent reliable over longer, more complex work takes more: loops that check its output, start a new run each time an event occurs, and help the agent improve from past runs.
1:309K views · 24d ago · 1:30
The longer an agent works, the more repeated, contradictory, or stale information builds up in its context. Models perform worse as their context grows, a problem Jeff Huber and his team at Chroma named Context Rot. In our new course, Context Engineering, taught by Jeff Huber alongside guest speaker Drew Breunig, you'll build agent harnesses from scratch in Python and learn to give agents the right information, in the right format, at the right time.
1:3319K views · 1mo ago · 1:33
You'll learn foundational data engineering skills while implementing modern data architectures using open-source tools Work through the full data engineering lifecycle stages: generating, ingesting, storing, transforming, and serving data Build batch and streaming pipelines, then orchestrate, automate, and monitor data flow Model, transform, and serve data for analytics and machine learning
2:1412K views · 1mo ago · 2:14
Guest: On-Device Memory
A cloud-based assistant can remember what it sees, hears, and is told, but only by staying connected and sending that data to a remote server. In this course, you'll learn how to Build a local memory system that stores, retrieves, filters, and forgets text and image memories. Build an offline assistant that recalls experiences across text, voice, and images. Teach an AI application to recognize new objects from stored examples without retraining a model.
1:01:34410K views · 1mo ago · 1:01:34
Guest: Coding Agents
Vibe coding is fast, but it often produces code that doesn't match what you asked for. Spec-driven development is the disciplined alternative: write a clear markdown spec defining what to build, and let your coding agent implement it.
1:383K views · 2mo ago · 1:38
A note from Paul Everitt for everyone taking the course: local AI is just getting started, and there’s room for you to shape it. Take the course, then come share what you build
3:3815K views · 2mo ago · 3:38
Guest: AI Dev
Thousands of developers, two days on the San Francisco waterfront, and a whole lot of code. Sessions covered the problems developers are actually solving right now: agentic systems, context engineering, multimodal applications, reliability in production, and building with AI coding agents. Speakers came from Google DeepMind, LangChain, Oracle, AMD, Arm, Neo4j, Replit, LandingAI, and AI21 Labs.
2:0811K views · 2mo ago · 2:08
Introducing Building Adaptive AI Agents, a short course built in partnership with Oracle and taught by Nacho Martínez, Data Scientist Advocate at Oracle, and Casius Lee, AI Developer Advocate at Oracle. Agents don't need to relearn the job every session: give them mechanisms to carry experience forward, and they get faster and more reliable with every run.
1:575K views · 3mo ago · 1:57
AI coding agents like Claude Code and Codex give you a ready-made way to work. That convenience comes at the cost of control: over which models run, what you spend, and what leaves your machine. In our new short course, AI Coding Workflows: From Cloud to Local, built in partnership with JetBrains and taught by Paul Everitt, Developer Advocate at JetBrains, you'll take that control back one layer at a time.
2:105K views · 3mo ago · 2:10
In this finalist demo for the 7-Day Voice AI Builder Challenge with @VocalBridgeAI, developer Sapna shows how voice escalation prevents data loss during database cleanup. The Scenario When deduplicating customer accounts for Acme Corp, the AI agent finds an old record with active orders conflicting with a new empty import. Since deletion is irreversible, the agent refuses to guess. Instead, it calls Sapna for vocal confirmation before executing the cleanup.
Topics: Task setup in Claude Code (TASK.md execution) · Agent detects duplicate records & identifies permanent deletion risk · Irreversible action gate triggers @VocalBridge voice escalation · Inbound phone call detailing the account conflict (2021 vs. 2024) · Developer selects 2021 record; AI asks for explicit confirmation · Call ends & agent executes exact database deletion script (cleanup.py)
5:121K views · 3mo ago · 5:12
In this finalist demo for the 7-Day Voice AI Builder Challenge (in collaboration with @VocalBridgeAI) developer Eugenia sets an autonomous AI agent to work on product launch positioning while she goes AFK for a walk, instructing it to call her only if it hits a decision only a human can make!
Topics: Setting the AFK instruction ("going for a walk, call me if needed") · Autonomous handling of routine positioning tasks · Escalation matrix triggered at a high-stakes decision point · Outbound phone call via @VocalBridge pitches strategy options · Interactive voice Q&A (developer asks about pricing risks) · Decision locked in ("Lead on Value") with a 7-day review task added
6:174K views · 3mo ago · 6:17
In this 1st place demo for the 7-Day Voice AI Builder Challenge with @VocalBridgeAI developer Nick sets an AI agent to fix a checkout bug AFK, calling his phone only when a decision requires human judgment. The Scenario While working on the checkout API code, the AI coding agent hits a critical decision point and needs to decide on a fix Option A: Maintain backward compatibility (safer, duplicates code).
Topics: AI Agent initiates outbound phone call to developer · Language fallback (Greek to English handling) · AI presents technical trade-offs (Option A vs Option B) · Developer makes decision via voice · Agent automatically executes Option B, logs the choice, and updates code
1:404K views · 3mo ago · 1:40
Learn how to make AI code review effective and build your own review agent in this short course, "AI Code Review," built in collaboration with Qodo and taught by Nnenna Ndukwe. The quality of an AI code review depends more on the context the reviewer has than on the model behind it.
2:193K views · 4mo ago · 2:19
When an LLM generates text, much of the time goes to moving the model's weights from memory to the compute units, not to the math itself. On a GPU those weights sit off-chip, and a large model may be split across several chips, so data travels back and forth before computation can even start.
2:285K views · 4mo ago · 2:28
Voice is one of the most natural human interfaces, but adding it to AI applications has historically forced a tradeoff: fast voice-to-voice models that sacrifice reliability, or accurate speech-to-text-to-LLM-to-speech pipelines that add latency. This course teaches you how to get both, using Vocal Bridge's architecture that pairs a real-time foreground agent with a reasoning background agent.
2:128K views · 5mo ago · 2:12
Introducing Fast & Efficient LLM Inference with vLLM, a short course built in partnership with Red Hat and taught by Cedric Clyburn, Senior Developer Advocate at Red Hat. Serving open-source LLMs efficiently, for many users at low latency and reasonable cost, comes down mostly to memory management. Two things compete for that memory: the model weights and the KV cache.
25:58171K views · 5mo ago · 25:58
Guest: Andrew Ng
If you’ve never written code before, this course is for you. In less than 30 minutes, you’ll learn to describe an idea in words and let AI transform it into an app for you. You’ll build a working web application in minutes: A funny interactive birthday message generator that runs in your browser and can be shared with friends. Then you’ll customize it by telling AI how you want it changed, tweaking it until it works exactly how you want.
24:3712K views · 5mo ago · 24:37
This talk by Cline's Ara Khan explains why they went from "evals are useless" to using them as a core part of my agent improvement loop. I share practical heuristics for interpreting, running, and creating evals, and why doing them anyway is better than pure "vibes".
14:201K views · 5mo ago · 14:20
AI agents fail in unpredictable ways that traditional testing can't catch — hallucinations, wrong tool calls, policy violations, and more. Teams only discover these failures after users hit them in production. A simulation sandbox gives you a controlled environment with realistic users, tools, and workflows where you can run hundreds of scenarios against your agent before it ships, catching edge cases and adversarial inputs that would be impossible to test manually.
26:001K views · 5mo ago · 26:00
Modern enterprises don't struggle to experiment with AI — they struggle to operationalize it reliably. In this talk, CrewAI's CEO outlines how leading organizations are moving beyond one-off automations to build recurring, governed, and deeply embedded workflows that drive real business outcomes. Drawing on lessons from production deployments, João explores how to design systems that are auditable, scalable, and aligned with enterprise controls — without sacrificing speed.
19:181K views · 5mo ago · 19:18
From centralized to distributed: In the old world, organizations relied on one centralized data and AI platform. In the new world of AI agents, every agent needs its own sandboxed, secure, and modern data stack. In this 20-minute talk with live demo by Spice AI's Luke Kim, he explores why this architectural shift is critical and the key patterns required to give agents reliable, real-time data.
25:27851 views · 5mo ago · 25:27
The next major shift in enterprise AI is underway; enterprises are moving from generic AI they rent to specialized AI they own. The benefits are clear: higher quality, dramatically lower costs, full control, and a quality improvement flywheel while in production.
30:24618 views · 5mo ago · 30:24
At AI Dev 26 x San Francisco, Flower Lab's Daniel Beutel talked about Flower SuperGrid, the industry standard for Federated AI. With SuperGrid Agents, you can now build and run context-rich agents that learn from interactions, access sensitive data and (soon) collaborate with other SuperGrid Agents.
18:44627 views · 5mo ago · 18:44
Most agentic systems rely on hardcoded heuristics to navigate execution decisions (e.g. which models, tools, and test-time compute scaling approaches to use) leading to efficiency leakage across cost, latency and accuracy.
17:23607 views · 5mo ago · 17:23
In this talk by Zencoder's Andrew Filev, attendees learned how decomposing tasks into pipelines and dynamically routing them across models improves quality, reduces cost, and makes AI systems more reliable.
26:24740 views · 5mo ago · 26:24
Building your first agent is exciting. Building a platform that can evolve into an office where dozens of teams can safely deploy their own agents is a different beast entirely.
28:17998 views · 5mo ago · 28:17
More code, fewer staff — the industry is on a bender. But what about quality? At AI Dev 26 x San Francisco, Paul Everitt from JetBrains discussed the rise of agentic engineering and how old lessons can be adapted to build new professional practices.
14:04520 views · 5mo ago · 14:04
This talk by Qdrant's Thierry Damiba shows how to build a real-time video anomaly detection system that works in open-world settings, where the most important events are often the ones you did not explicitly train for. It walks through an edge-to-cloud architecture using Qdrant Edge, Twelve Labs, and NVIDIA Metropolis to detect unusual behavior, search video semantically, and support grounded investigation workflows.
25:251K views · 5mo ago · 25:25
Every AI coding tool can generate code. Very few can generate the right code for your organization, because they're missing context. They don't know why your team chose Redis over DynamoDB, what the team decided in a Slack thread from two months ago about the auth migration, or which architectural patterns your principal engineers actually enforce in review.
31:01652 views · 5mo ago · 31:01
The future of automating knowledge work depends on AI agents that can reliably read and understand documents — but today's agents struggle with complex layouts, tables, and visual elements.
16:06761 views · 5mo ago · 16:06
Voice AI today is mostly customer service bots. That's about to change — and AI devs will build what comes next. This talk by Vocal Bridge's Ashwyn Sharma introduces Voice UI as an emerging interface category, explains the technical architecture that makes truly multimodal voice experiences possible, and shows you exactly how to build one live on stage.
17:39712 views · 5mo ago · 17:39
Guest: AG UI
All UI will be AI. In this talk, Atai Barkai from CopilotKit explored the emerging space of Generative UI, and went over the practical building blocks emerging in the space.
16:21364 views · 5mo ago · 16:21
As computer-use models become smarter, the bottleneck for their adoption is becoming clear: 20 years of web identity tech must be securely adapted for AI agents. In this talk, Idan Raman from Anchor Browser explored how enterprises and builders are solving this challenge.
36:18823 views · 5mo ago · 36:18
We all know how fast it is to create an app with modern AI agents but how do we ensure the code is reliable, maintainable and secure enough to be used by enterprises? In this talk, Sonar's Tom Howlett shared a benchmark from their testing of 35 (and growing) of the latest and highest-performing large language models and showed how they compare not just on task completion but on the quality of the code they create.
30:23455 views · 5mo ago · 30:23
AI coding tools have dramatically increased developer velocity, but they haven't eliminated the need for code review. In fact, they've made it more critical than ever. In this session, Erik Thorelli, Developer Experience Lead at CodeRabbit, shares how engineering organizations can operationalize AI-driven code review as a scalable, production-grade quality gate.
Showing the 36 most recent videos. Everything plays right here on The Pod Atlas.
Have thoughts about DeepLearningAI? Share them with other learners.
Write a review← Back to all educational channels