Table of Contents

E-commerce content Generation.

Social Media Content Management

Website Optimization

E-commerce content generation.

Creating engaging content is a great way to drive traffic to your e-commerce platform. Multimodal models can help you create content for your web platform, such as blog posts or a news section, products and service descriptions, and page content. They can also help you create products or services. Ideally, AI content should be a starting point and not an end destination.

Automated Content Creation

In practice, you can automate content creation on any platform, like social media channels, applications, or your website, using APIS from your platform and the multimodal AI models you want to use. Every major cloud AI player, such as Azure, Google Cloud, and Amazon, has a cloud run function option that allows you to run code directly in a container with minimal configuration, enabling immediate execution. In Google Cloud, this is called Cloud Run, which lets you run containerized applications with automatic scaling. In Amazon Web Services (AWS), it’s called AWS Lambda, offering serverless code execution without managing infrastructure. In Microsoft Azure, it’s known as Azure functions.

Blog Post Generator

We will use these functions to run some python code to create an application for automatic blog post creation.

Step1: Enable WordPress API.

  1. Log in to WordPress as an administrator.
  2. Navigate to Users > Your Profile.
  3. Scroll down to the Application Passwords section.
  4. Create a new application password by providing a name.
  5. Copy the generated password, as it will be used in your Google Cloud Run function for API authentication.

You will use this application password along with your WordPress REST API URL to post content programmatically. Your url will look like this: https://yoursite.com/wp-json/wp/v2/. Replace yoursite.com, with the domain name of your website.

Step 2: Start a new project on your Cloud platform.

Visit the Google Cloud Console and sign in with your Google account.

Create a New Project:

  • Click on the project dropdown at the top of the page.
  • Select an existing project or click New Project to create one.
  • Provide a project name and billing account.

Step 3: Enable the APIS on Google Cloud

In the left-hand menu, click APIs & Services > Library. In the API Library, type the API name you want to enable in the search bar. You need to activate the Cloud Run API (to manage and deploy your containerized app), Cloud Build API (to build the app into a container), Artifact Registry API (to store container images), Gemini for Google Cloud API (for generating AI-based content), Cloud Logging API (to monitor and debug the app), and IAM API (to manage permissions and access control). Once these APIs are enabled, you can use them within your project to automate tasks and run your application.

Step 4: Create an API Key in Google AI Studio

Go to aistudio.google.com and click on “Generate a New API Key.” Select the Google project you just created and generate the key. Copy the key to use it in your Google Cloud project.

Step 5: Create a service account

We must create a service account in Google IAM & Admin to ensure your cloud function runs. Locate IAM & Admin in the left-hand menu of the Google Cloud Console. Under IAM & Admin, click on Service Accounts. This will open the section to view existing service accounts or create a new one.

Once you click on Create New, you will see the following screen. Add a name and a description. Afterward, the wizard will ask you to assign a role. For Google Cloud Run, the most appropriate role to assign to the service account is Cloud Run Admin. This role allows the service account to fully manage Cloud Run services, including creating, updating, and deleting services. Grant this service account access to the project section; click the Select a role dropdown. Search for Cloud Run Admin or find it under By product or service > Cloud Run. Select Cloud Run Admin.

Step 6: Go to Cloud Run and Deploy your code.

First, access the Google Cloud Console and navigate to Cloud Run. Once there, click Create Service to begin. Choose the option inline editor, choose the region us-east1 , and the runtime Python 3.12

After you click Create, you must add your Python code to the editor. Click on edit and delete the code that is already there. The essential files you’ll need are your main.py (which contains your Python application logic) and requirements.txt (which lists the necessary dependencies). The main.py file includes the code to interact with APIs or perform the required tasks, and the requirements.txt file will list libraries like Flask, requests, and any others your application requires.

In this notebook, you will find the code you need to copy and paste to the main.py file:

https://colab.research.google.com/drive/1-uy92SeXB4d6PLRUZ4Bu12CwXy8sdJGY

in the requirements.txt file add the following dependencies: 

functions-framework==3.*
requests
Flask
Pillow
google-generativeai>=0.3.0

These dependencies are needed because your code uses Flask to create a web service that handles incoming requests, requests to communicate with external APIs like WordPress and Gemini, Pillow to process any images involved, google-generative to generate content using AI, and functions-framework to ensure the app runs smoothly in Google Cloud Run without manual server management.

During deployment, Cloud Run will automatically build your application from these files and prepare it for use. You’ll also be asked to configure the entry point—in this case, it will likely be the main function within your Python file, which Cloud Run invokes when the service is accessed.

Next, configure any environment variables your application needs, such as API keys or credentials. Your application will interact with external services like the Gemini API and WordPress, so you should add environment variables such as GEMINI_API_KEY, WORDPRESS_USERNAME, and WORDPRESS_APP_PASSWORD. We already copied the GEMINI_API_KEY in step 4. You can find the WORDPRESS_USERNAME, and WORDPRESS_APP_PASSWORD by going to the dashboard of your website, go to users, and get the username and password associated with your username.

Once you’ve finished configuring the settings, click Create. Google Cloud Run will then build and deploy your application, making it accessible at a unique URL. This URL will allow you to interact with your deployed Python service via HTTP requests. Cloud Run will automatically handle scaling and performance based on demand.

Step 7: Go to WordPress and Create the PHP code to display the blog post generator.

Go to your website and install a plugin called Snippets.

This plugin will allow you to add code snippets to your website. Once this is installed, you can add a new php snippet. You will find this code in the notebook:

https://colab.research.google.com/drive/1R4Q5_9x-MVSgajNhfB26YVMi5i677zm_#scrollTo=mvgKVwNLAWCQ

The php code will create a form we can insert in any page using a shortcode [docs_ai_post_generator].

AI Post Generator

Enter a basic topic - AI will create an engaging title
Provide basic details - AI will expand this into a full blog post

Image and Content Generators

Midjourney

Midjourney is an AI research lab based in California. In 2022, it launched the Midjourney application inside of Discord. The free tool was available to anyone. Midjourney quickly became popular, reaching 16 million members by October 2023. Due to its fast-growing user base, Midjourney canceled its free usage options, and now it can only be accessed via a subscription or APIs. Midjourney is one of the highest-performing realistic text-to-image generators on the market. Known for its versatile and high-quality outputs, the company became one of the leaders in AI image generation.

Stability.ai

Stability.ai is a UK-based AI unicorn focused mainly on generative artificial intelligence. Created in 2019, the company became famous for its Stable Diffusion AI models. Stable Diffusion models are largely open-source models that help users generate images from text and create 3D objects, video, and audio. The model outputs from Stability.ai can be edited in various ways, including image enhancement, creating new images from the same style, painting pictures, removing backgrounds, upscaling an image, erasing specific details, etc. The Stability.ai models can be accessed for free via Hugging Face’s library.

Other image generation tools include amongst others: DALL·E (Open AI), Gemini , Adobe Firefly.

Perplexity. ai

Perplexity.ai was created in 2022 in California. The company made a free-to-use search engine focused on directly answering questions based on internet research. Users can ask follow-up questions, search for related images or videos, and curate news based on their interests. The search engine includes sources and helps you organize queries.

Jasper

Jasper was created in 2021 in Austin, Texas. It was developed to assist businesses and content creators in generating high-quality content using AI. Jasper helps to write blog posts, social media posts, product descriptions etc..

Video Creation and Editing Platforms.

Videos are among the most engaging forms of content across all social media and the internet. In video content development, AI models make editing processes easier and shorten production timelines. While advanced video generation capability is at an infant stage, multi-modal models are good enough to generate short videos from text prompts, images, and audio inputs. These models enable creators to create high-quality videos faster and with less manual intervention than ever.

With text-to-video tools, machine learning algorithms first analyze the text using natural language processing techniques to understand the context and extract the main elements. Then, the model renders the scenes for the description by assembling the video frame by frame. Advanced techniques, including those from General Adversarial Networks (GANs), ensure the frames generated are realistic, while tools like Blender animate the scenes. OpenAI’s SORA and Phenaki have already shown this, making creating exciting video content with a few lines of text much more effortless. Image-to-video systems analyze still images to detect motion patterns and inject missing frames into a smooth, continuous video. Classic algorithms, including motion estimation and optical flow, predict how objects should move from frame to frame to support consistency throughout. Users can quickly transform static photos into dynamic video sequences thanks to platforms like Runway and Simplified. Audio-to-video generation uses the power of multimodal models with state-of-the-art audio inputs to generate corresponding visuals. The AI will synchronize animations to either music or speech, generating elements in a video that align with the mood or rhythm of the audio.

Other domains in which AI models excel include video-to-video translation. These models can turn a day scene into a night scene and change a cartoon into a realistic animation, among other things. The AI smooths out the visual properties of the video with either CycleGAN or diffusion models for smooth inter-frame transitions. Vid2Vid from NVIDIA enables creators to apply various stylistic transformations to their videos.

AI can also be used to create deepfake videos. In this method, the algorithms learn from face or object datasets by analyzing them and then inserting those elements in other videos. AI models achieve this by using facial mimicking, expressions, and motion, complemented by voice synchronization. DeepFaceLab and FaceSwap are two industry-leading platforms to create deepfakes.

Veed.io

Veed.io is an online video editing platform designed to make video creation simple and accessible for users of all levels. It operates directly in the browser, eliminating the need to download software. The platform offers editing features, including trimming, cropping, adding text, overlays, transitions, and music. With its AI-powered tools, users can automatically generate subtitles, remove backgrounds, clone voices, insert avatars, and enhance videos.

Adobe Express

Adobe Express enables you to create and edit videos directly in your web browser without additional software installations. The platform offers a variety of tools, including cropping, trimming, splitting scenes, and adding animated effects. You can add text, icons, and music to enhance your videos. Adobe Express provides numerous templates to help you get started quickly and supports direct sharing to social media platforms.

BU offers a free account for students.

Sora.AI (Open AI)

Sora is OpenAI’s large-scale generative model designed to create high-quality videos from text prompts. It uses a diffusion transformer architecture that generates videos of varying durations, resolutions, and aspect ratios, up to one minute long.

Sora can extend videos, perform advanced editing, or create videos from scratch using text prompts. Sora learns to generate videos from text by training on large amounts of video data, each paired with a description. During this process, it learns to associate words in the text (like dog or running) with visual patterns in the video. To make learning more manageable, Sora breaks down videos into small pieces called patches and knows how these patches change over time to create motion. When given a text prompt, Sora starts by generating a rough, noisy video version, then gradually refines it, adding details that match the text. A transformer model helps maintain consistency across the frames, ensuring that objects and movements align correctly throughout the video. Through this process, Sora can create realistic videos from simple text descriptions. Sora is not publicly available yet but will be available this year. Read the documentation on Open Ai’s website to learn more about how Sora works.

Runway

Runway is an applied AI research company based in New York. The company develops generative tools and artificial intelligence systems that enable users to create videos, do advanced video editing, and create audio and 3D textures. Runway just launched its Gen-3 Alpha foundation model, which was trained on videos and images. Gen-3 Alpha powers advanced editing features like Motion Brush ( to create motion for objects), Advanced Camera Controls, and Director Mode while introducing new tools for more precise control over structure, style, and motion in videos.

CogVideo

Researchers from Tsinghua University in Beijing developed a powerful AI model called CogVideo. This model takes text descriptions and turns them into high-quality videos. CogVideo can generate dynamic, full-motion videos based on simple text prompts. CogVideo’s ability to transform detailed text into coherent video sequences makes it unique. A video artist using Cog even won a Cannes Jury Award for short films for his film “The Crow.”

Try out the Cog Model on Hugging Face.

Social Media Content Management.

Social media content management using AI has become increasingly popular and sophisticated. AI tools can help create engaging and personalized experiences at scale, saving time and resources. Most social media platforms already integrate AI tools to give advertisers advanced insights into their audiences, create content faster, and maximize return on investments. Natural Language Processing-based tools can help develop posts, captions, and even responses to user interaction on your posts. On the other hand, AI-powered video and image generation software develops visually stunning content from simple inputs.

Beyond creation, AI provides real-time insight into social media performance: it enables marketers to analyze user behavior, understand sentiment, and find trends or patterns in content consumption and sharing. These insights drive businesses to make data-informed decisions and iterate on strategies to better adapt to their audience.

AI also plays a critical role in making social media safer and more trustworthy. Machine learning models help detect and remove fake accounts, spam, and inappropriate content to prevent platform abuse. In addition, various AI-powered systems have been developed to spot and combat so-called fake news by monitoring misinformation and warning users about possible risks. However, AI is also a two-edged sword because it aids in catching fake accounts and misinformation. At the same time, it has enabled deepfakes and synthetic content, complicating the fight against disinformation. Bad actors use AI to fabricate fake news stories, images, and videos to deceive audiences, further adding to new ethical challenges in social media.

Native Platforms

Meta

Meta launched Advantage Plus on their ad platform in 2022. Via Advantage +, marketers can automatically create up to 150 different versions of advertisements by just uploading a few media assets. Ads are personalized, based on users’ demographics, intents, history, and interests, and tailored to perform well in diverse placements such as, for example, on Instagram Reels or Facebook Stories. AI assists with campaign creation, audience segmentation, choosing ad placements, and maximizing exposure for the amount paid. Moreover, Meta integrated the Meta AI assistant on its platforms to create posts and images quickly. Meta AI can also be a free, alone, standing chatbot and an alternative to other AI bots such as GPT-4 or Gemini. Meta recently added voice-control and multimodal options to Meta AI. The Meta AI bot is not as refined as its competitors. Still, it can help users learn more about their favorite social media content quickly or instantly generate visuals to share with their network. Meta also gives users free access to an AI studio to create AI characters to engage with other users or represent themselves as creators. This helps to engage users or autorespond to particular inquiries.

Tiktok

TikTok has a set of artificial intelligence tools similar to Meta. TikTok features Symphony Assistant and Symphony Creative Studio, which creators can use to automate content creation. The TikTok Symphony Assistant helps you generate ideas and scripts for videos, provides insights, or showcases trends on the platform.

Symphony Creative Studio goes one step further, allowing you to generate videos from a few product or page links. The studio allows for editing and refining these AI-generated videos before publishing. It also offers translation and dubbing features to adapt content for diverse audiences. A forthcoming feature currently in testing will allow custom-created digital avatars as narrators. Right now, users can select from a database of AI-generated avatars based on real creators.

Third-Party Tools

Sprout Social

Sprout Social is an all-in-one social media management platform that integrates with social media channels like Instagram, Facebook, TikTok, LinkedIn, Threads, etc. It offers various options, from social media scheduling to AI-driven sentiment analysis, to help brands discover what customers say about their companies. It also uses machine learning to discover content trends and copy for social media posts. It has advanced performance tracking and publishing tools. The company offers a free 30-day trial.

SocialBee

Social Bee is a versatile social media creation platform that integrates with YouTube, LinkedIn, Instagram, Facebook, X, Pinterest, and other social media channels. The platform provides a copilot AI solution for developing social media strategy ideas, scheduling posts, categorizing content, monitoring performance, generating content, and delivering content ideas.

https://socialbee.com/ai-post-generator/

Hubspot

Hubspot is one the leading customer relationship management (CRM) systems in the world. Created in 2006 by two MIT students, in Massachusetts, it caters mostly to small and medium sized businesses. Since 2023, Hubspot integrated AI tools into its platform. For social media management, Hubspot helps to schedule posts, or generate content.

Website Optimization

Hotjar

Hotjar is a behavior analytics and feedback tool that uses features like heatmaps, session recordings, and surveys to help you understand how users interact with your site. AI is used to analyze user behavior patterns and generate insights.

Semrush

Semrush, founded in 2008, is an online visibility management platform that enables businesses to run search engine optimization, pay-per-click, content, social media, and competitive research campaigns to improve their online marketing results. The company has integrated AI across many of its tools, including the SEO Writing Assistant, ContentShake AI, Semrush Copilot, and AI-powered features in its Listing Management Tool, to enhance content creation, provide SEO insights, and automate various aspects of digital marketing.