AI Dev Tools Zoomcamp: AI-Native Software Engineering

A free, hands-on course on using AI developer tools to build, test, deploy, extend, and audit software without losing engineering discipline.

Register | Course Platform | Join Slack | Course Slack Channel | Telegram | YouTube Playlist | FAQ

Quick Links

Resource	Link
Register	Sign up for the free 2026 cohort
Start date	August 31, 2026
Learn	Course platform · Materials · Videos
Course launch	Read the launch article
Community	Slack · Course channel · Telegram
Help	Docs · Logistics · FAQ

[!NOTE] The 2026 materials are currently being finalized. The repository shows the direction for the next cohort, but some videos, homework, deadlines, project requirements, and module details may change before the cohort starts on August 31, 2026.

About the Course

AI Dev Tools Zoomcamp is a free, practical course for developers and technical data professionals who want to learn disciplined AI-assisted software development.

The course is built around one core workflow:

Give AI tools the right context, use them for the right job, review what they produce, test the result, and ship software with guardrails.

You will compare modern AI developer tools, build and deploy a full-stack application, operate it with observability and an agent-assisted incident response, and extend coding agents with MCP and reusable capabilities.

This is not a prompt-engineering-only course, and it is not a course about training models or building RAG systems. The focus is day-to-day software development with AI coding assistants, agents, tests, CI/CD, deployment, documentation, and review.

Who Should Join

This course is for you if you:

• Already write basic code in Python, JavaScript, TypeScript, or a similar language.
• Want to use AI coding tools professionally, not only for isolated snippets.
• Want a repeatable workflow for planning, prompting, reviewing, testing, debugging, and shipping code.
• Care about maintainability, tests, deployment, documentation, and code review.
• Want to build a project that demonstrates practical AI-assisted development skills.
• Benefit from cohort deadlines, community support, peer review, and certificate eligibility.

The course is especially relevant for software engineers, ML engineers, MLOps engineers, AI engineers, data scientists, data engineers, analysts who write code, and technical students who want practical project experience.

This course is probably not the right fit if you have never programmed before, want a non-technical AI introduction, or mainly want model training, fine-tuning, LangChain, vector databases, or RAG.

Prerequisites

You do not need prior AI tools experience.

You should have:

• Basic programming ability in Python, JavaScript, TypeScript, or a similar language.
• Comfort using the command line.
• Git and GitHub basics.

Helpful, but not required:

• Basic web development familiarity.
• Basic Docker familiarity.

Not required:

• Prior coding assistant or coding agent experience.
• A powerful machine or GPU.

How to Take the Course

You can follow the course as part of the live cohort or use the materials self-paced.

	Live Cohort	Self-Paced
Start	August 31, 2026	Anytime
Cost	Free	Free
Lectures	Pre-recorded	Pre-recorded
Homework	Graded	Available, but not scored
Leaderboard	Yes	No
Peer review	Yes	No
Certificate eligibility	Yes	No
Registration	Sign up here	Start from the repository

[!IMPORTANT] "Live cohort" does not mean mandatory live classes. Lectures are pre-recorded. Live cohort means shared deadlines, scored homework, leaderboard participation, peer review, community momentum, and certificate eligibility.

Self-paced steps:

1. Follow the materials in this repository.
2. Watch the videos in the course playlist.
3. Ask questions and share progress in DataTalks.Club Slack.
4. Do the homework for practice and build a portfolio project.

Syllabus

Module 1: AI-Native Developer Workflow

Learn how to turn a vague product idea into specified, implemented, and independently verified software with coding agents.

In this module, you'll:

• Turn a vague idea into a spec and a backlog of focused tasks.
• Give coding agents durable context through AGENTS.md.
• Use PM/engineer/QA roles with loop and graph engineering to implement and verify the backlog.

Outcome: design and run an AI-native development workflow that takes a product from idea to implementation and independent QA.

Module 2: Build and Ship an AI-Assisted Full-Stack App

Build a full-stack application with AI coding assistants, replacing mocked components one at a time.

You will:

• Turn an idea into a spec, then build a frontend prototype and an OpenAPI contract.
• Implement a FastAPI backend from the contract, then connect it to the frontend with auth and real-time collaboration.
• Replace temporary storage with SQLite and add tests.

Outcome: a working full-stack app with real-time updates, persistent data, tests, and a clear API contract.

Module 3: Test, Containerize, and Deploy an AI-Assisted App

Take the application from your machine to a public deployment, with tests and delivery checks that verify the whole system.

You will:

• Containerize the app and move from SQLite to Postgres with Docker Compose.
• Add integration and end-to-end tests, including Playwright for the collaborative workflow.
• Deploy to AWS and wire up a GitHub Actions CI/CD pipeline.

Outcome: a public, containerized app backed by Postgres, with automated tests and CI/CD that deploys only after tests pass.

Module 4: DevOps and Observability for AI-Built Apps

A deployed app can still fail silently. Add release controls and observability so you can detect, investigate, and respond to problems.

You will:

• Separate development and production environments with a promoted, versioned release pipeline.
• Instrument the app with OpenTelemetry into Prometheus, Loki, Tempo, and Grafana, and set up an actionable alert.
• Give a coding agent a bounded, read-only responder role to investigate incidents.

Outcome: separate dev/prod workflows, an observability stack, actionable alerts, and a proof-of-concept AI on-call responder.

Module 5: Coding Agent Building Blocks: Reusable Skills and Specialized Subagents

Turn repeated workflows and project roles into reusable capabilities that coding agents can discover and apply.

You will:

• Create discoverable skills with SKILL.md and decide whether they should be global or project-specific.
• Define focused subagents, including PM/engineer/QA roles, to separate implementation from independent review.
• Orchestrate tasks in parallel using isolated Git worktrees.

Outcome: reusable skills, subagent definitions, and an orchestration pattern for sequential or parallel project work.

[!NOTE] Module 5 has no graded homework. It is assessed through the module deliverable instead.

Final Project

The final project applies the course workflow to an end-to-end application of your own.

You will:

• Choose a problem and describe the expected system behavior
• Build a frontend
• Build a backend
• Define frontend/backend communication with an OpenAPI contract
• Add database or persistent storage
• Test the application
• Containerize the application
• Deploy it so others can try it
• Document how AI tools, prompts, workflows, agent instructions, MCP, or automation were used

For the live cohort certificate path, projects are evaluated through peer review. Learners must pass the final project and complete the required peer reviews.

Certificate

Certificates are available during the live cohort for learners who complete the final project and the required peer reviews.

Homework helps you stay on track, but the certificate is based on the final project and peer review requirements. See Certification for the general certificate process.

Self-paced learners can use all materials for free, but certificate eligibility requires participation in a live cohort.

Instructors

• Alexey Grigorev
• Bhavani Ravi

Learner Stories

This course fundamentally changed how I approach AI development. I moved from "building models" to designing AI-assisted systems that are faster to ship and easier to iterate on.

During the course, I built a portfolio optimization tool, a full-stack application using ChatGPT, Lovable, and Antigravity, and a structured GitHub project with reproducible documentation.

• Yann Pham-Van, Freelance Data Scientist

The course taught me how to use coding agents effectively, debug issues, and gave me exposure to MCPs, tools, and prompts. It helped me conceptualize any idea into a working prototype.

• Revathy Ramalingam, Senior Software Engineer at Yalabs Solutions

During the course I built a Finnish learning website which helps English users practice reading, writing, listening, and speaking skills for the Finnish language.

I used Antigravity with Gemini and Claude, a Context7 documentation MCP server, TypeScript, Python, Next.js, FastAPI, SQLite, and CI/CD with GitHub Actions.

• Kaiquan Mah, Data Scientist at Total eBiz Solutions

Community and Support

Slack

Join DataTalks.Club Slack and use the #course-ai-dev-tools-zoomcamp channel for course questions, troubleshooting, and discussion.

Before posting questions, please check:

• Question guidelines
• Community guidelines

Announcements

Course announcements are posted in the Telegram channel. Telegram messages are also reposted to the course Slack channel.

Learning in Public

Share your progress with the hashtag #aidevtools and tag Alexey Grigorev or DataTalksClub. Learning in public helps you stay accountable, meet other learners, and build visible proof of your work.

See the learning in public guide.

Sponsors

Interested in supporting DataTalks.Club and this course? Reach out to alexey@datatalks.club.

FAQ

For the full list, see the AI Dev Tools Zoomcamp FAQ.

Is this course free?

Yes. The videos, materials, homework, and live cohort participation are free.

Do I need AI experience?

No. Prior AI coding tool experience is not required. You should be able to write basic code and work with a project repository.

Do I need to be a software engineer?

No. The course is also relevant for data scientists, ML engineers, MLOps engineers, data engineers, analysts who write code, and technical students. You do need basic programming ability.

Are there live classes?

Lectures are pre-recorded. The live cohort adds deadlines, scored homework, a leaderboard, peer review, community support, and certificate eligibility.

Can I take the course self-paced?

Yes. You can use the videos and repository materials anytime. Self-paced learners are not certificate-eligible unless they join a live cohort.

Do I need a powerful computer or GPU?

No. A powerful machine or GPU is not required.

Is this course about RAG or model training?

No. The course focuses on AI-assisted software development workflows: coding assistants, agents, MCP, testing, CI/CD, deployment, documentation, review, security, audit, and DevOps.

About DataTalks.Club

DataTalks.Club is a global online community of data enthusiasts. It is a place to discuss data, learn, share knowledge, ask and answer questions, and support each other.

Website | Join Slack Community | Newsletter | Upcoming Events | YouTube | GitHub | LinkedIn | X

Most DataTalks.Club activity happens in Slack. We post updates there, discuss course questions, and help learners support each other.

At DataTalks.Club, we organize online events, community activities, and free courses. Learn more in the DataTalks.Club docs.