Starts 1 June 2026 · 10 spots $500 · 4 weeks · Tashkent

Ship production AI features in 4 weeks

A flagship course for working backend, frontend, and mobile developers who want AI skills now — not in a year. Curated stack, practitioner instructor, and a deployed AI service you can show on your resume.

Apply for Next Cohort View Curriculum

10 working devs per cohort · offline campus, Uzbek-language instruction

agent_loop.py 
# Week 3 — agents, no framework magic
def run(query, tools, max_steps=8):
    msgs = [user_msg(query)]
    for _ in range(max_steps):
        r = llm.chat(msgs, tools=tools)
        if not r.tool_calls:
            return r.content
        for call in r.tool_calls:
            result = tools[call.name](call.args)
            msgs += [r, tool_msg(call.id, result)]
    raise StepBudgetExceeded

Deployed

FastAPI + Docker

Week 4

Demo Day

Free content is everywhere. Most of it is broken.

Every dev team needs at least one person who can build with AI.

Tutorial chaos

50 contradictory how-tos, each picking one tool with no explanation. You drown instead of build.

Outdated within months

80% of free LLM tutorials break within a year. Deprecated models, changed SDKs, frameworks now considered anti-patterns.

No one reviews your code

A Colab notebook, never a deployed service. No senior ever sees your code. You can't tell good from bad.

What you can't get free

6 things this course gives you that YouTube can't

AI chatbots answer questions, but they can't replace cohort, mentorship, or production judgment.

01

Curated production stack

One opinionated stack — and the reasoning behind every choice. Skip 2-3 months of tool paralysis.

02

Cost & latency engineering

Token economics, prompt caching, batching, self-hosting trade-offs. The stuff free content skips.

03

Guaranteed-current 2026 stack

80% of free LLM tutorials break within a year. Ours runs in 2026 — with deprecations flagged.

04

Live code review

Senior eyes on your actual project, every week. The one thing free content can't deliver.

05

A deployed AI service

Not a notebook. Not a localhost demo. A real, containerized service on the internet — a hireable signal.

06

Cohort of 10 working devs

Peer learning + a contact list of AI-capable Tashkent devs to refer, hire, or collaborate with.

Common objection

"Why not just use ChatGPT to learn?"

“

AI tutors only answer questions you know to ask. Beginners don't know the right questions.

“

AI doesn't hold you accountable to ship anything. Cohort, deadlines, and demo day do.

“

AI can't compress a senior engineer's years of judgment into 4 weeks. A practitioner instructor can.

Sukhrob Nuraliev - AI Bootcamp Instructor

Practitioner Instructor

Working AI engineer

Meet Your Instructor

Sukhrob Nuraliev

I'm a working AI engineer, not a full-time educator. I'll teach you what I actually use to ship production AI features — the curated stack, the cost trade-offs, the things that break in prod and how to fix them.

Direct code review on your weekly project — not just a graded checkbox

Live Q&A on the messy real-world questions free content avoids

Offline campus in Tashkent, Uzbek-language instruction

Learn With Me

4 weeks. From LLM API call to deployed AI service.

Every week ships a project. The course ends with a deployable portfolio piece on the internet.

Is this course for you?

Honest answer in 10 bullets.

You're a great fit if:

  • You want a deployed portfolio piece, not another certificate
  • You want to ship AI features, not study ML theory
  • You're a working dev (backend, frontend, mobile) or a CS student who already codes
  • You can commit 10-15 hours per week for 4 weeks
  • You can read code in any language — Python helps but isn't required

This isn't for you if:

  • You've never written code before
  • You're hoping to become an "AI Engineer" in 30 days — we explicitly don't promise that
  • You want deep ML research (training models from scratch, papers, math)
  • You prefer self-paced video courses
  • You're already shipping AI features in production

One price. No surprises.

Full upfront payment keeps the cohort serious and the focus on shipping.

4-Week Flagship Course
$500 one-time

Full upfront payment for the complete 4-week course in Tashkent.

What's included:

  • 4 weeks of hands-on, project-based instruction
  • Live teaching from a working AI engineer (Sukhrob Nuraliev)
  • A weekly project shipped — CLI tool, RAG system, agent, deployed service
  • Live code review on your final project
  • A deployed AI service on the internet, with a writeup
  • Cohort of 10 working devs — peer learning & network
  • Demo day to present your work
Apply for the Next Cohort
10 spots per cohort · Light application screening

Questions about the course? Check our FAQ or apply and ask

Frequently asked questions

Everything you need to know about the course.

Apply for the next 4-week cohort.

Starts 1 June 2026. 10 spots. Tashkent. Light screening — we want serious working devs who'll show up and ship.

Apply Now

Limited spots. We'll reach out within 48 hours.

@
+998 ( ) - -

By applying, you agree to be contacted about the program.