AI LAUNCH LAB — LIVE 30-DAY COHORT

From Prompt to Product in 30 Days

Build the brand. Build the product. Build the system. One real project taken from idea to brand, website, application, campaign, automation — and launch. Live, hands-on, in a batch of 20–25 builders.

Early Bird — enroll by July 21 and pay 7,888 HKD instead of 12,888 HKD.

TWO WORLDS, ONE COURSE

Create demand. Build the product. Converge on one launch.

TRACK 01 — MARKETING & CREATIVE

Create demand

The systems that generate attention — taught by Ryan.

  • A1Brand and art direction
  • A2AI images and campaigns
  • A3Advertisements and short-form content
  • A4Cinematic AI video
  • A5Voice, avatars, dubbing, and localization
  • A6Growth and launch campaigns

TRACK 02 — TECHNOLOGY & SYSTEMS

Build the product

The systems that turn ideas into working software — taught by Daksh.

  • B1Websites
  • B2Applications and games
  • B3Multimodal AI
  • B4MCP integrations
  • B5Agents and subagents
  • B6Automation, infrastructure, and VPS deployment

ONE FUNCTIONING, MARKETABLE LAUNCH

The tracks alternate all month. Every creative asset feeds the build; every build ships with a campaign. They are different systems — and they converge on Day 30.

WHAT STUDENTS BUILD

One project. Every layer connected.

Nothing here is an isolated demo. Each output feeds the next — one connected evolution from idea to launch.

01IDENTITY

  • Brand system
  • Campaign visuals

02PRODUCT

  • Live website
  • Application or game
  • Multimodal product

03STORY

  • Cinematic advertisement
  • Voice / avatar content

04SYSTEMS

  • MCP integration
  • Deployed agent

05LAUNCH

  • Launch campaign
  • Portfolio case study

These are course aims — the outputs serious students work toward across 30 days and present at Demo Day. Commercial results, income, or client acquisition are not guaranteed.

MEET THE INSTRUCTORS

Two instructors. Two complementary skill sets. One complete launch system.

Learn the creative systems that generate attention from Ryan, and the technical systems that turn ideas into functioning products from Daksh.

Ryan Chi, Marketing and Creative Instructor, speaking on stage
RYAN CHITRACK 01

MARKETING & CREATIVE INSTRUCTOR

Ryan Chi

AGE 22 · 7 YEARS EXPERIENCE

Ryan teaches the creative and growth side of the course: how to position an offer, build a distinctive brand, produce AI-generated campaigns, direct cinematic video, create multilingual voice and avatar content, and turn creative output into a repeatable launch system.

  • President, 0xU Club
  • Member, Penn Blockchain
  • Founding Team, City Protocol (Jump Invested)
  • CMO, Academic Labs (HTX, Signum, UOB Invested)
  • Guest Lecturer, CUHK (#18 QS WORLD RANKING) and HKU EMBA

COURSE FOCUS

  • AI marketing strategy
  • Brand creation and art direction
  • AI-generated images and campaigns
  • Short-form ads and content systems
  • Cinematic AI video production
  • ElevenLabs voice, dubbing, localization
  • AI avatars and spokesperson content
  • Launch campaigns and creative services

Ryan creates attention, persuasion, visual identity, and market demand.

Daksh builds the product, integrations, infrastructure, and automation.

Together they teach the complete path from idea to functioning, marketable launch.

Daksh Joshi, Technology and Systems Instructor, presenting on stage
DAKSH JOSHITRACK 02

TECHNOLOGY & SYSTEMS INSTRUCTOR

Daksh Joshi

AGE 23 · 7 YEARS EXPERIENCE

Daksh teaches the product and systems side of the course: how to use coding agents to build websites, applications, games, MCP integrations, multimodal products, and production automations that can be deployed and maintained on a VPS.

  • Founding Team and CPO, Public AI ($10M Raised)
  • Technical BD, Asatralane (Solana Node)
  • Member and Head of Tech, 0xU Club
  • Guest Lecturer, CUHK (#18 QS WORLD RANKING)

COURSE FOCUS

  • AI-assisted website and app development
  • Claude Code, Codex, terminal workflows
  • MCP servers and integrations
  • Context engineering, long-running projects
  • APIs, databases, multimodal applications
  • Agents, subagents, and automation
  • VPS, Linux, Docker or PM2, domains, HTTPS
  • Games, internal tools, AI-powered products

EXPERIENCE ACROSS

The Chinese University of Hong KongThe University of Hong KongWharton — University of PennsylvaniaPublicAIStanford UniversityInsiders.bot

Factual references to the instructors' individual experience and affiliations. They do not imply sponsorship or endorsement of AI Launch Lab by any institution or company named.

THE COURSE, IN ONE VIDEO

Watch how one idea becomes a launch.

THE COMPLETE 30-DAY CURRICULUM

Four weeks. Twelve classes. One Demo Day.

Marketing and Creative classes alternate with Technology and Systems classes all month. Every live class runs the same rhythm: 15 min concepts · 45 min instructor live build · 35 min student implementation · 15 min critique · 10 min assignment handoff.

MARKETING & CREATIVE — RYANTECHNOLOGY & SYSTEMS — DAKSH

WEEK 1

Foundations, Branding, and Websites

CLASS 01The AI Operating SystemTECHNOLOGY & SYSTEMS

TOPICS

  • Choosing models for research, writing, coding, and creative work
  • Prompt structure, constraints, examples, and reference materials
  • Hallucinations and verification
  • ChatGPT, Claude, Perplexity, and NotebookLM workflows
  • Context windows and persistent project memory
  • Long-running project management
  • When to begin a new session
  • Master briefs, decision logs, task files, and handoff documents
  • File and folder structures
  • Privacy, sensitive data, and API-key safety

STUDENT OUTPUT

  • A complete project context pack
  • Master brief
  • Product definition
  • Audience file
  • Decision log
  • Reusable prompts
  • Session handoff document
CLASS 02Build a Complete Brand Using AIMARKETING & CREATIVE

TOPICS

  • Customer personas and positioning
  • Naming and tagline development
  • Brand personality and tone
  • Mood boards and reference gathering
  • Image prompting: lighting, lens, composition, materials, camera angle, and colour treatment
  • Product and character consistency
  • Style references
  • Inpainting, outpainting, and image editing
  • Campaign graphics and social templates
  • Building a reusable brand kit

CORE TOOLS

GPT ImageMidjourneyAdobe FireflyCanvaClaude Design or equivalent

STUDENT OUTPUT

  • Mini brand book
  • Mood board
  • Visual direction
  • Ten consistent campaign images
  • Reusable image-prompt system
CLASS 03Build and Publish a Website Using AITECHNOLOGY & SYSTEMS

TOPICS

  • Product requirements
  • Wireframes and information architecture
  • AI-generated interface concepts
  • Frontend development
  • Forms and basic functionality
  • Databases and authentication
  • Mobile responsiveness
  • GitHub and version control
  • Deployment and domains
  • Analytics and error monitoring

CORE TOOLS

Claude Designv0LovableReplitBoltClaude Code or CursorSupabaseGitHubVercel or VPS deployment

STUDENT OUTPUT

  • A responsive live website
  • Working contact or registration flow
  • Connected repository
  • Deployment documentation

WEEK 2

Advertising, Applications, and Cinematic Video

CLASS 04AI Content and Advertisement EngineMARKETING & CREATIVE

TOPICS

  • Researching customer pain points
  • Hooks and angles
  • Hook–problem–solution–proof–CTA structure
  • Static advertisements
  • Short-form advertisements
  • UGC-style scripts
  • Founder-led content
  • Product demonstrations
  • Carousels and thumbnails
  • Producing multiple variations
  • Repurposing long-form content
  • A/B testing
  • Brand consistency
  • Content calendars

CORE TOOLS

ChatGPT or ClaudeGPT Image or MidjourneyCanvaCapCutDescriptElevenLabs

STUDENT OUTPUT

  • Three advertisement concepts
  • Five social posts
  • Seven-day content calendar
  • Reusable content-production workflow
CLASS 05Build an AI Application or GameTECHNOLOGY & SYSTEMS

TOPICS

  • Requirements and user flows
  • Frontend and backend concepts
  • APIs
  • Databases
  • Authentication
  • State management
  • Debugging with screenshots and logs
  • Planning before coding
  • Testing AI-generated code
  • Git commits and version restoration
  • Environment variables
  • Deployment

POSSIBLE PROJECTS

  • Quiz game
  • AI character game
  • Recommendation application
  • Proposal generator
  • AI marketing-audit tool
  • Lead qualification application
  • Portfolio generator
  • Browser game with leaderboard

CORE TOOLS

Claude CodeCodexCursorReplitv0BoltLovableSupabase

STUDENT OUTPUT

  • A functioning public mini-product or game
  • Test feedback from real users
  • Basic documentation
CLASS 06Create a Cinematic AI AdvertisementMARKETING & CREATIVE

TOPICS

  • Concept development
  • Visual references
  • Character and product sheets
  • Shot lists and storyboards
  • Establishing, medium, and close-up shots
  • Lens choices
  • Camera movement
  • Lighting continuity
  • Product and character consistency
  • First-frame and last-frame control
  • Image-to-video versus text-to-video
  • Generating separate shots
  • Editing a coherent narrative
  • Sound design
  • Music and voiceover
  • Colour grading
  • Export formats

CORE TOOLS

HiggsfieldRunwayLumaPikaElevenLabsSunoCapCut or Premiere

STUDENT OUTPUT

  • A coherent 20–45 second cinematic advertisement
  • At least five generated shots
  • Final edited version with audio

WEEK 3

Multimodal Products, Voice, Avatars, and MCP

CLASS 07Build a Multimodal AI ApplicationTECHNOLOGY & SYSTEMS

TOPICS

  • Multimodal AI
  • Transcription versus video understanding
  • Semantic search
  • Embeddings
  • Video indexing
  • Structured extraction
  • Timestamped search
  • API calls
  • JSON responses
  • Retrieval-augmented generation
  • Cost and latency controls

EXAMPLE PROJECT

  • Upload a video
  • Search for specific moments
  • Ask questions about it
  • Generate a summary
  • Produce timestamped results
  • Generate a narrated output

CORE TOOLS

TwelveLabsElevenLabsClaude or OpenAI APIsSupabaseReplit or Claude Code

STUDENT OUTPUT

  • A multimodal application
  • One custom feature such as chapters, advertisement analysis, clip recommendations, or searchable moments
CLASS 08Voice Cloning, Avatars, and Multilingual ContentMARKETING & CREATIVE

TOPICS

  • Voice prompting and emotional direction
  • Voice cloning with explicit consent
  • Voiceovers
  • AI sound effects
  • Dialogue cleanup
  • Dubbing
  • Translation versus localization
  • Tone consistency between languages
  • Lip sync
  • AI avatars
  • Personal digital twins
  • Recurring spokesperson characters
  • Disclosure, consent, and likeness rights

CORE TOOLS

ElevenLabsHeyGenSynthesiaDescriptCapCutSuno

STUDENT OUTPUT

  • One spokesperson video
  • One additional language version
  • Voice and localization workflow
CLASS 09MCPs: Connect AI to Real Tools and DataTECHNOLOGY & SYSTEMS

TOPICS

  • MCP hosts, clients, and servers
  • Tools, resources, and prompts
  • Local and remote MCP servers
  • Connecting existing servers
  • Authentication
  • Read-only and write access
  • User approval
  • Wrapping APIs
  • Testing and inspection
  • Logging and error handling
  • Prompt-injection risks
  • TypeScript or Python MCP implementation

POSSIBLE PROJECTS

  • Asset library MCP
  • Notion knowledge MCP
  • Google Drive search MCP
  • Lead database MCP
  • Product catalogue MCP
  • Analytics MCP
  • Content calendar MCP

STUDENT OUTPUT

  • One safe MCP server
  • At least two tools
  • At least one resource
  • Basic access-control and usage documentation

WEEK 4

Growth, Agents, Infrastructure, and Monetization

CLASS 10Launch and Grow an AI-Created ProductMARKETING & CREATIVE

TOPICS

  • Market and competitor research
  • High-intent audiences
  • Offer creation
  • Landing-page messaging
  • Lead magnets
  • Email sequences
  • Organic distribution
  • Paid advertisement structure
  • Influencer and creator outreach
  • Content calendars
  • Retargeting concepts
  • Conversion tracking
  • Creative performance
  • Moving from attention to leads and sales

STUDENT OUTPUT

  • 30-day content calendar
  • Three advertisement concepts
  • One lead magnet
  • Five-message email or DM sequence
  • One landing-page variation
  • Basic measurement dashboard
CLASS 11Agents, Subagents, Automation, and VPS DeploymentTECHNOLOGY & SYSTEMS

TOPICS

  • Chatbots versus workflows versus agents
  • When not to use an agent
  • Deterministic steps versus model decisions
  • Specialized subagents
  • Planner, researcher, builder, and reviewer roles
  • Context passing
  • Skills, hooks, and reusable instructions
  • Human approval gates
  • Rate limits and budgets
  • Retries and failure handling
  • Schedules and webhooks
  • Logs and observability
  • SSH and Linux basics
  • VPS setup
  • Environment variables and secrets
  • PM2 or Docker
  • Backups
  • Domains and HTTPS
  • Keeping services online

POSSIBLE WORKFLOW — LEAD OPERATIONS

  • New lead
  • Company research
  • Lead scoring
  • Personalized proposal
  • Human approval
  • Gmail draft
  • CRM update

ALTERNATIVE WORKFLOW — CONTENT OPERATIONS

  • Content idea
  • Research agent
  • Script agent
  • Visual prompt agent
  • Brand-review agent
  • Human approval
  • Scheduling

CORE TOOLS

Claude CodeCodexClaude subagentsn8nVPSDocker or PM2

STUDENT OUTPUT

  • One deployed automation or agent workflow
  • Human approval stage
  • Logging
  • Restart and maintenance documentation
CLASS 12Turn AI Skills Into a Service or BusinessMARKETING & CREATIVE

TOPICS

  • Choosing a niche
  • Selling outcomes rather than "AI"
  • Productized services
  • Before-and-after case studies
  • Portfolio construction
  • Proposal writing
  • Scope and revision limits
  • Client onboarding
  • Credential and asset collection
  • Value-based pricing
  • Maintenance retainers
  • Handover documentation
  • Demonstration videos
  • Responsible AI disclosures

POSSIBLE OFFERS

  • AI advertisement studio
  • Video localization service
  • Website and automation package
  • AI content engine
  • Internal AI assistant
  • Lead-generation workflow
  • Product-photography package
  • Avatar-content service
  • MCP integration service

STUDENT OUTPUT

  • Commercial offer
  • Portfolio case study
  • Proposal
  • Final presentation

DAY 30 — JOINT DEMO DAY

Every student presents a complete launch.

  1. 01The problem
  2. 02The target customer
  3. 03Brand and creative direction
  4. 04Live website
  5. 05Application or game
  6. 06Cinematic advertisement
  7. 07Agent, automation, or MCP
  8. 08Launch strategy
  9. 09What they would improve next

SUGGESTED ASSESSMENT

Working product or application25%
Marketing and commercial usefulness20%
Creative quality and consistency20%
Automation or technical implementation20%
Presentation, documentation, and responsible use15%

Ready to build all of it?

July 21 – August 19, 2026 · Only 20–25 students · Enrollment on WhatsApp

Join the July 21 Batch on WhatsApp

TOOLS COVERED

Organized by outcome, not by logo.

Every tool below appears in the curriculum in service of something you ship. Tool names are used factually as tools covered — no partnership or sponsorship is implied.

RESEARCH AND REASONING

ChatGPTClaudePerplexityNotebookLM

IMAGES AND DESIGN

GPT ImageMidjourneyAdobe FireflyCanvaClaude Design

VIDEO

HiggsfieldRunwayLumaPika

VOICE, MUSIC, AND AVATARS

ElevenLabsSunoHeyGenSynthesia

WEBSITES AND APPLICATIONS

Claude CodeCodexCursorv0LovableReplitBoltSupabaseGitHub

MCP, AGENTS, AND AUTOMATION

MCPn8nSubagentsAPIs

MULTIMODAL UNDERSTANDING

TwelveLabs

INFRASTRUCTURE

VPSDocker or PM2

ENROLLMENT

One batch. One month. One conversation to join.

FORMAT
30-day live cohort
DATES
July 21 – August 19, 2026
RHYTHM
Three live classes per week
CLASSES
12 core classes + joint Demo Day
CAPACITY
Only 20–25 students
METHOD
Project-based learning

Included: all 12 live classes, the practical output built in every class, the complete idea-to-launch curriculum journey, and the Day 30 joint Demo Day. Nothing else is promised here — ask anything on WhatsApp.

PRICE

7,888 HKD

12,888 HKD

EARLY BIRD — JULY 21 BATCH

Payment is arranged directly with the team on WhatsApp.

Join the July 21 Batch on WhatsApp

There is no online checkout. Enrollment and payment are discussed and completed directly in a WhatsApp conversation with the team.

FAQ

Straight answers.

Is coding experience required?

No. Coding experience is useful but not required for the introductory builds. The course is designed for motivated beginners and intermediate users, and every technical class is a guided live workshop.

Is this live or recorded?

The cohort is live: three live, project-based classes per week for 30 days, ending in a joint Demo Day on Day 30.

What tools do I need?

A laptop and the tools covered in the curriculum. Tools are organized by outcome on this page — research and reasoning, images and design, video, voice and avatars, websites and applications, MCP and agents, multimodal understanding, and infrastructure.

Are software subscriptions included?

No. Students purchase their own tool subscriptions, so every account, workspace, and asset is personally yours — and you keep using them after the course ends.

What will I build?

One real project, end to end: a brand system, a live website, an application or game, campaign visuals, a cinematic advertisement, voice or avatar content, a multimodal product, an MCP integration, a deployed agent, a launch campaign, and a portfolio case study — presented at Demo Day.

How large is the batch?

20–25 students. The batch is deliberately small so every project gets live critique and troubleshooting time in class.

How do I enroll?

Message the team on WhatsApp. Enrollment and payment are discussed and completed directly in that conversation — there is no online checkout.

What happens on Demo Day?

On Day 30, every student presents their complete launch: the problem, the target customer, brand and creative direction, the live website, the application or game, the cinematic advertisement, their agent, automation, or MCP, the launch strategy, and what they would improve next.

Is income or employment guaranteed?

No. The course teaches skills and produces real artifacts, but income, employment, clients, virality, and business success are not guaranteed.

AI LAUNCH LAB

One idea. Thirty days. A complete launch.

30-DAY LIVE COHORTSTARTS JULY 21, 2026ONLY 20–25 STUDENTS
WhatsApp — July 21 Batch