Practical AI Marketing Applications
From idea to an automated AI system
8 sessions for marketers, business owners, freelancers and newcomers to the field. Learners bring a laptop, pick a real brand from Session 1 and build a real campaign throughout the course: setting up an AI assistant loaded with the brand standards, planning content, producing designs and videos, running multi-channel ads, building a customer care chatbot and assembling everything into one operating system.
What challenges do you face when bringing AI into your marketing work?
The 6 most common situations for marketers and small business owners.
Many AI tools, no idea where to start
A new tool appears every week. It is unclear which tool does what, and you install plenty but none of the jobs runs end to end.
AI-written content feels lifeless, like every other brand
Without the brand standards loaded as context, the output is generic and off-tone, and needs almost a full rewrite, which takes longer than writing it yourself.
You have content but cannot produce images and video yourself
You depend on designers and freelancers for every post. You want to post reels and TikTok consistently, but production is always the bottleneck.
Running ads without knowing who to target, reports you cannot read
Reports with dozens of columns, but no idea which group to turn off and which to give more budget. You decide by gut feeling, then wonder where the money went.
An inbox flooded with repeated questions, and slow replies lose customers
Dozens of questions about prices, opening hours and how to order repeat every day. It takes staff to cover them, yet messages outside hours are still missed.
You know each part but have not assembled them into one system
Content, design, video, ads and customer care are each done in their own way. There is no operating schedule, and you cannot measure how many hours AI saves.
A hands-on course: leave class with a working AI system you can use right away
No AI theory for its own sake. Every session is done directly on the learner's real brand, and the output of one session is the input of the next.
The output of one session is the input of the next. Session 2 produces the content plan, image prompts and video scripts; Sessions 3 and 4 turn them into designs and videos; Sessions 5 and 6 take them into ads; Session 7 builds the chatbot; Session 8 assembles the system. Do the early work well and the later sessions get easier.
Each learner picks a brand from Session 1 and brings a logo, product photos and a price list. Everything produced in class is an asset you can use right after the course, not a hypothetical exercise.
A principle throughout: every request to AI starts with positioning, voice and visual standards. This is the line between content with identity and content anyone could make.
The quality of Vietnamese content and voice varies hugely between tools. Each session states clearly which tools give usable Vietnamese and which do not.
When connecting AI to ad accounts, grant read-only access to reports, never permission to spend money or change campaigns. AI drafts; people review and are responsible for prices, product information and images.
Who is this course for?
3 core learner groups, and who is not yet a good fit.
In-house marketers
You handle content, design, ads or customer care and want to boost productivity with AI instead of doing every task by hand.
Small business owners and freelancers
You do your own marketing or manage a very thin team, and need a system that runs content, ads and customer care steadily without inflating outsourcing costs.
Newcomers who want to enter marketing
No technical or coding knowledge needed, just basic computer and social media skills. You leave with real work to show employers or clients.
You cannot bring a laptop to class
The whole course is hands-on, working directly on the machine at the pace of the lecture. Without a laptop, it is almost impossible to follow.
You have not chosen a brand to work on throughout the course
You need a real brand with a logo, product photos and a price list. Working on an imaginary brand means the output cannot be reused.
You want to learn AI engineering, models and programming
The course teaches applying AI to everyday marketing work. It does not cover model architecture, fine-tuning or programming.
You expect AI to fully replace people
The course principle is that machines draft, people review and take responsibility. Never let AI finalize prices, promotions or publish unverified content.
An 8-session roadmap: one production line
Session 2 is the brain that produces the plan, image prompts and video scripts; the following sessions turn them into designs, videos, ads, a chatbot and finally a complete working system.
Phase 1: Foundations & Content Production
FOUNDATION: Sessions 1, 2, 3 & 4
Session 1
AI Foundations & Account SetupTool map · Setup · Prompts · Brand AI assistant · 3 hours
- An overview of AI tools for marketing: what AI can and cannot do; a tool map by task: planning and content, design, video, ads, customer care
- A tool cost table: which need a paid plan, which free versions are enough for learning, and which plan to buy
- Hands-on setup of each account with illustrated guides; language settings, two-factor security, safe payment methods; the assistant cross-checks every table
- Core skill: the structure of a good prompt, including role, context, request, output format and examples; basic markdown to present prompts clearly
- Main output: build a brand AI assistant: create a dedicated project and load positioning, voice, visual standards and product information; test it with a trial writing task to confirm the voice is right
Session 2
Building a Marketing Plan & Content Plan with AIThe brain session: plan, content and prompts for the later sessions · 3 hours
- A 4-level line of thinking: business goal → marketing plan → content plan → detail of each post
- Practice building a monthly marketing plan from the brand's real goals using the AI assistant built in Session 1
- Allocate 5 content pillars in proportions that fit your industry; schedule posts by day and hour
- Write each post in detail with the right mood and tone; editing technique: ask AI to revise in a specific direction instead of rewriting from scratch
- Generate image prompts for Session 3 and detailed video scripts for Session 4, the input material for the entire second half of the course
Session 3
AI cho DesignProduct photos by guideline · Designs on Weavy · Templates · 3 hours
- Product imagery principles: composition, lighting, color, props; the line between AI images and real photos
- Create images from the prompt library of Session 2; how to refine prompts to get the right composition and brand color tone
- Mandatory verification: AI-generated product images must reflect reality; an image that looks better than the real product is false advertising
- Produce designs on Weavy combined with Figma: set up frames by logo, color, fonts and safe zones; export multiple sizes for each channel
- Main output: a reusable template set for the 3 most common design formats, so others on the team can reuse them while keeping brand standards
Session 4
AI cho VideoScript to video · Voiceover · CapCut · Batch editing · 3 hours
- The structure of an effective short video: hook, content, call to action
- Build videos from the script library of Session 2 using AI tools; turn product photos into video
- Add Vietnamese voiceover with ElevenLabs and add subtitles
- Edit in CapCut by guideline: pacing cuts, music, subtitle fonts, logo, brand frame; note music copyright and each platform's format
- Main output: a batch video editing workflow: edit many videos from many scripts in one session, with a reusable preset and template set
Phase 2: Distribution & Operating System
GROWTH: Sessions 5, 6, 7 & 8
Session 5
AI + Facebook / Instagram AdsCustomer research · Ad variations · Reading reports with AI · 3 hours
- Two purposes for running ads: building awareness and collecting customer data; which metrics measure each
- Use AI to sketch customer personas and find insights from the brand description and existing data
- Create many ad content variations from one core message; an A/B test framework: change only one element at a time, and conclude only with enough data
- Highlight of the session: connect a connector that pulls live reports into Claude and ask for analysis in Vietnamese; a fallback layer exports reports to a file for anyone without account access, with the same results
- Platform policy and safety: rules on AI content, avoiding account suspension, verifying prices and promotions before running
Session 6
AI + Google / TikTok AdsSearch & Google Business · TikTok Ads · Channel comparison · 3 hours
- Google when customers are actively searching; for brands with stores, Google Business and Maps are a must-optimize point
- Use AI to write the Google Business description, suggest keywords customers often search and how to reply to reviews
- TikTok Ads: platform specifics, reuse the videos made in Session 4, analyze and optimize campaigns with AI
- Main output: compare cost per result across Facebook, Google and TikTok on real data, and produce a budget allocation proposal for next month
Session 7
AI for Customer Care & ChatbotsCare scripts · Zalo OA · Facebook · Testing · 3 hours
- The questions customers message most: price, stock status, delivery, opening hours, how to order, product advice
- Safety boundaries: when to let AI reply automatically, when to hand over to a real person; the chatbot never finalizes prices or promises what the brand cannot deliver
- Build automatic replies on Zalo OA: write a set of answers in the brand voice and scripts that lead to closing an order
- Build automatic replies for the Fanpage inbox and sync scripts with Zalo; scripts to keep customers leaving their information
- Test with real situations, fix wrong answers and hand over operating rules to the team
Session 8
The Overall AI Operating Process: CapstoneAssemble the system · Safety · Measure value · Present · 3 hours
- Combine the 5 areas learned (content, design, video, ads, customer care) into one system with the brand AI assistant at the center and a reviewer at the end
- Weekly operating schedule: when to produce content, when to run ads, when to review the chatbot; a clear split between human tasks and machine tasks
- Three AI risks and verification rules before publishing: wrong information and prices, images that differ from reality, leaked customer data
- Measure value: calculate the hours the system saves each week and estimate the outsourcing costs it replaces, figures to take back in your report
- Capstone: each learner presents the AI system and campaign they built, and the instructor reviews and adjusts each case on the spot
8 deliverables to take home and use right away
All materials, prompt sets and templates are handed over in editable formats.
Brand AI assistant
An AI project loaded with positioning, voice, visual standards and product information, used for every later task instead of describing everything from scratch each time.
Monthly content plan + post details
A content plan allocated by pillar and posting schedule, with details of at least 10 posts in the brand voice.
Image prompt library & video scripts
A library of image prompts by guideline and detailed video scripts, reusable for the following months.
Design set + design templates
Designs for the posts in the content plan and a reusable template set for the 3 most frequent design formats.
Short videos + batch editing workflow
Videos for the content plan with presets, templates and a batch production workflow to keep a steady posting rhythm.
Ad variation set + test plan
Ad content variations with different emotional angles and lengths, with an A/B test framework and a report analysis prompt set.
Zalo OA & Facebook chatbot script set
A set of automatic replies in the brand voice, order closing scripts and rules for when to hand over to a real person.
AI system + value measurement sheet
An operating diagram, weekly schedule, safety checklist and a sheet measuring hours saved and the outsourcing costs replaced.
Hands-on Learning Experience
Learn by doing on your own brand, not by taking home slides to read.
A short lecture, then work right away on your own brand
Each session runs in 20-30 minute cycles: a short lecture, learners work on that part immediately, a quick review, then the next part. No long theory blocks that cause overload.
Someone goes table to table to solve problems
The class always has a technical assistant to help with setup and account and tool errors, plus backup demo accounts so nobody gets stuck mid-lesson.
Ask for analysis in Vietnamese right in class
Connect a connector that pulls live reports into Claude, ask which group to turn off and which to give more budget, and require it to cite the exact columns so you can check them against the original file.
Present your system and get direct feedback
In Session 8, each learner presents the AI system and campaign built for their brand; the instructors and assistants review and adjust each case on the spot.
Course instructors
Three instructors lead three areas: strategy and content, AI production, and multi-channel advertising.
Tien Tien
Founder AAU Academy & PEACH Consulting
Leads the brand positioning part and the thread from business goals down to the content plan, ensuring everything AI produces ties to the learner's real business problem.
Nguyen Ngoc Tuan
Founder & CEO Transform Group
Guides the application of AI to content production, design and video: from prompts based on brand standards to batch editing workflows.
Tran Quan
Founder & CEO Admatrix
Leads the multi-channel advertising block: customer research with AI, creating ad variations, connecting a connector to read reports and comparing performance across Facebook, Google and TikTok.
Tuition and what's included
Editable materials handed over
All 32 documents, guides and prompt sets are numbered TL-01 to TL-32, handed over as editable files so learners can customize them for their own brand.
Two-tier options for the ads section
Those with account access connect a connector to read live reports; those without export reports to a file and analyze them. The lesson works for every learner.
Tuition does not include the cost of paid accounts for AI tools; the tool and cost table is sent before the course starts so you can register in advance. Free versions can still be used for learning but are limited. The class stays small, with at least 1 technical assistant helping at each table.
Frequently asked questions
Answers to common questions before you register.
Can I learn if I know nothing about AI?
Yes. The course requires no technical or coding knowledge, just basic computer and social media skills. Session 1 is fully dedicated to account setup and building the AI assistant, with an assistant checking each table.
What should I prepare before the course starts?
A laptop, tool accounts already registered per the cost table sent in advance, and a real brand to work on throughout the course, with a logo, product photos and a price list. If you already have brand standards, bring them to load into AI in Session 1; if not, Session 1 guides you through quickly building a minimal version.
Do I have to buy paid tools?
Not required. Free versions are enough for learning but are limited in usage and features; we recommend buying the paid plan of the main tools to practice fully. A detailed cost table is sent before the course starts.
What if I have no ad account or no access?
Sessions 5 and 6 always have two options: connect a connector to read live reports for those with access, or export reports to a file and analyze them with AI, with the same results. Both options are read-only, with no permission to spend money or change campaigns.
Is it safe to connect AI to ad accounts?
The course rule is to grant read-only access to reports when connecting to an account with billing attached, never share passwords, and use the platform's official permission settings. Writing and changing campaigns is only demonstrated on the instructor's account.
Can I reuse the materials after the course?
Yes. All prompt sets, design and video templates, chatbot scripts and forms are handed over in editable formats so learners can customize them and hand them over to their own team.
When is the next intake?
The course starts on Friday 18 Sep 2026, with evening classes on Monday, Wednesday and Friday from 18:00 to 21:00 in Ho Chi Minh City, ending on 05 Oct 2026. See the full schedule on the Course Schedule page or call 0565 66 76 86 to reserve a seat.
Start the AI Marketing system for your brand
Leave your details and the AAU team will contact you for a free consultation within 24 hours and confirm the nearest intake schedule.
Or call now 0565 66 76 86 (Mon–Sat, 9h–18h)