Tips, Expertise, Articles and Advice from the Pro's for Your Website or Blog to Succeed
Those with a keen eye will notice that the up and down arrows for the number input doesn’t show up in Chrome (and some other WebKit based browsers) until you hover or are focused on the input. To force the arrows to be visible we just need to change the opacity of the inner and outer spin button.
The post Force Up and Down Arrow for Number Input appeared first on Web Design Weekly.
WordPress shortcodes are a simple way to set up functions to create macro codes for use in post content. For instance, the following shortcode (in the post/page content) would add your recent posts into the page: It’s pretty simple and brings your WordPress blog alive with ease. Recent Post Short Code In WordPress 1 Add this code to your functions.php file. 2 … Read more
The post Recent Post Shortcode appeared first on Web Design Weekly.
This snippet is super small but is super handy. It detects to see if JavaScript is enabled on the users browser. If yes, a class will be added to the <body> tag Adds the class yourClassName to <body> tag if Javascript is enabled.
The post Add Class To Body If JavaScript Is Enabled appeared first on Web Design Weekly.
Large language models (LLMS) are trained to predict the next token (set of characters) following an input sequence of tokens. This makes LLMs suitable for unstructured textual responses.
However, we often need to extract structured information from unstructured text. With the Python LangChain module, you can extract structured information in a Python Pydantic object.
In this article, you will see how to extract structured information from news articles. You will extract the article's tone, type, country, title, and conclusion. You will also see how to extract structured information from single and multiple text documents.
So, let's begin without ado.
As always, we will first install and import the required libraries.
The script below installs the LangChain and LangChain OpenAI libraries. We will extract structured data from the news articles using the OpenAI GPT-4 latest LLM.
!pip install -U langchain
!pip install -qU langchain-openai
Next, we will import the required libraries in a Python application.
import pandas as pd
import os
from typing import List, Optional
from langchain_core.pydantic_v1 import BaseModel, Field
from langchain_openai import OpenAI
from langchain_openai import ChatOpenAI
from langchain_core.prompts import ChatPromptTemplate
We will extract structured information from the articles in the News Article with Summary dataset.
The following script imports the data into a Pandas DataFrame.
dataset = pd.read_excel(r"D:\Datasets\dataset.xlsx")
dataset.head(10)
Output:
To extract structured output, we need to define the attributes of the structured output. We will extract the article title, type, tone, country, and conclusion. Furthermore, we want to categorize the article types and tones into the following categories.
article_types = [
"informative",
"critical",
"opinion",
"explanatory",
"analytical",
"persuasive",
"narrative",
"investigative",
"feature",
"review",
"profile",
"how-to/guide",
]
article_tones = [
"aggressive",
"neutral",
"passive",
"formal",
"informal",
"humorous",
"serious",
"optimistic",
"pessimistic",
"sarcastic",
"enthusiastic",
"melancholic",
"objective",
"subjective",
"cautious",
"assertive",
"conciliatory",
"urgent"
]
Next, you must define a class inheriting from the Pytdantic BaseModel class. Inside the class, you define the attributes containing the structured information.
For example, in the following script, the title attribute contains a string type article title. The LLM will use the attribute description to extract information for this attribute from the article text.
We will extract the title, type, tone, country, and conclusion.
class ArticleInformation(BaseModel):
"""Information about a news paper article"""
title:str = Field(description= "This is the title of the article in less than 100 characters")
article_type: str = Field(description = f"The type of the artile. It can be one of the following : {article_types}")
tone: str = Field(description = f"The tone of the artile. It can be one of the following: {article_tones}")
country: str = Field(description= """The country which is at the center of discussion in the article.
Return global if the article is about the whole world.""")
conclusion: str = Field(description= "The conclusion of the article in less than 100 words.")
Next, you must define an LLM to extract structured information from the news article. In the following script, we will use the latest OpenAI GPT-4o LLM.
OPENAI_API_KEY = os.environ.get('OPENAI_API_KEY')
llm = ChatOpenAI(api_key = OPENAI_API_KEY ,
temperature = 0,
model_name = "gpt-4o-2024-08-06")
You need to define the prompt that instructs the LLM that he should act as an expert extraction algorithm while extracting structured outputs.
Subsequently, using the LangChain Expression Language, we will create a chain that passes the prompt to an LLM. Notice that here, we call the with_structured_output() method on the LLM object and pass it the ArticleInformation class to the schema attribute of the method. This ensures the output object contains attributes from the ArticleInformation class.
extraction_prompt = """
You are an expert extraction algorithm.
Only extract relevant information from the text.
If you do not know the value of an attribute asked to extract,
return null for the attribute's value."
"""
prompt = ChatPromptTemplate.from_messages([
("system", extraction_prompt),
("user", "{input}")
])
extraction_chain = prompt | llm.with_structured_output(schema = ArticleInformation)
Finally, we can call the invoke() function of the chain you just created and pass it the article text.
first_article = dataset["content"].iloc[0]
article_information = extraction_chain.invoke({"input":first_article})
print(article_information)
Output:
From the above output, you can see structured data extracted from the article.
In most cases, you will want to extract structured data from multiple text documents. To do so, you have two options: merge multiple documents into one document or iterate through multiple documents and extract structured data from each document.
You can merge multiple documents into a single document and then create a Pydantic class that contains a list of objects of the Pydantic class containing the structure data you want to extract. This approach is helpful if you have a small number of documents since merging multiple documents can result in the number of tokens greater than an LLM's context window.
To do so, we will create another Pydantic class with a list of objects from the initial Pydantic class containing structured data information.
For example, in the following script, we define the ArticleInfos class, which contains the articles list of the ArticleInformation class.
class ArticleInfos(BaseModel):
"""Extracted data about multiple articles."""
# Creates a model so that we can extract multiple entities.
articles: List[ArticleInformation]
Next, we will merge the first 10 documents from our dataset using the following script.
# Function to generate the formatted article
def format_articles(df, num_articles=10):
formatted_articles = ""
for i in range(min(num_articles, len(df))):
article_info = f"================================================================\n"
article_info += f"Article Number: {i+1}, {df.loc[i, 'author']}, {df.loc[i, 'date']}, {df.loc[i, 'year']}, {df.loc[i, 'month']}\n"
article_info += "================================================================\n"
article_info += f"{df.loc[i, 'content']}\n\n"
formatted_articles += article_info
return formatted_articles
# Get the formatted articles for the first 10
formatted_articles = format_articles(dataset, 10)
# Output the result
print(formatted_articles)
Output:
The above output shows one extensive document containing text from the first ten articles.
We will create a chain where the LLM uses the ArticleInfos class in the llm.with_structured_output() method.
Finally, we call the invoke() method and pass our document containing multiple articles, as shown in the following script.
If you print the articles attribute from the LLM response, you will see that it contains a list of structured items corresponding to each article.
extraction_chain = prompt | llm.with_structured_output(schema = ArticleInfos)
article_information = extraction_chain.invoke({"input":formatted_articles})
print(article_information.articles)
Output:
Using the script below, you can store the extracted information in a Pandas DataFrame.
# Converting the list of objects to a list of dictionaries
articles_data = [
{
"title": article.title,
"article_type": article.article_type,
"tone": article.tone,
"country": article.country,
"conclusion": article.conclusion
}
for article in article_information.articles
]
# Creating a DataFrame from the list of dictionaries
df = pd.DataFrame(articles_data)
df.head(10)
Output:
The above output shows the extracted article title, type, tone, country, and conclusion in a Pandas DataFrame.
The second option for extracting structured data from multiple documents is to simply iterate over each document and use the Pydantic structured class to extract structured information. I prefer this approach if I have a large number of documents.
The following script iterates through the first 10 documents in the dataset, extracts structured data from each document, and stores the extracted data in a list.
extraction_chain = prompt | llm.with_structured_output(schema = ArticleInformation)
articles_information_list = []
for index, row in dataset.tail(10).iterrows():
content_text = row['content']
article_information = extraction_chain.invoke({"input":content_text})
articles_information_list.append(article_information)
articles_information_list
Output:
Finally, we can convert the list of extracted data into a Pandas DataFrame using the following script.
# Converting the list of objects to a list of dictionaries
articles_data = [
{
"title": article.title,
"article_type": article.article_type,
"tone": article.tone,
"country": article.country,
"conclusion": article.conclusion
}
for article in articles_information_list
]
# Creating a DataFrame from the list of dictionaries
df = pd.DataFrame(articles_data)
# Displaying the DataFrame
df.head(10)
Output:
Extracting structured data from an LLM can be crucial, particularly for data engineering, preprocessing, analysis, and visualization tasks. In this article, you saw how to extract structured data using LLMs in LangChain, both from a single document and multiple documents.
If you have any feedback, please leave it in the comments section.
#701 — August 22, 2024
JavaScript Weekly
Regexes Got Good: The History (and Future) of Regular Expressions in JavaScript — Regular expression support was always a little underwhelming in JS, but things have improved. Steven takes us on a tour to refresh our knowledge, as well as show off his ‘regex’ library that boosts JS regexes to a true A++ rating. Steven was co-author of O’Reilly’s Regular Expressions Cookbook and High Performance JavaScript so knows his stuff.
Steven Levithan
WorkOS: The Modern Identity Platform for B2B SaaS — WorkOS is a modern identity platform for B2B SaaS, offering flexible and easy-to-use APIs to integrate SSO, SCIM, and RBAC in minutes instead of months. It’s trusted by hundreds of high-growth startups such as Perplexity, Vercel, Drata, and Webflow.
WorkOS sponsor
Node v22.7.0 (Current) Released — Node 22.6 let you strip types from source code, but now with –experimental-transform-types you can transform TypeScript-only syntax into JavaScript before running it too. Module syntax detection is now also enabled by default.
Rafael Gonzaga
Bun v1.1.25: Now Running at 1.29 Million Requests per Second — I’m having a little fun with the title, but the latest version of the JavaScriptCore-based JS runtime has added node:cluster support and uses this to demo a high level of HTTP throughput on a ‘Hello World’ example. Support for V8’s C++ API has also landed – notable because Bun isn’t V8-based.
Ashcon Partovi
IN BRIEF:
We’ve mentioned ECMASCript 2024 a bit recently, but Pawel Grzybek has a neat and tidy overview of what’s new in the ES2024 spec.
🐝 Could Wasp be ‘the JavaScript answer to Django’ for full-stack webdev? The Wasp team certainly thinks so.
🎙️ Ryan Dahl, creator of both Node.js and Deno, went on the Stack Overflow podcast to talk about Deno’s current limitations and what’s coming in Deno 2.0.
RELEASES:
PlayCanvas Engine 2.0 – A powerful JS-based Web graphics platform.
Node v20.17.0 (LTS) – The LTS release of Node adds support for require-ing synchronous ESM graphs.
Astro 4.14 – The popular agnostic content site framework now includes an experimental API for managing site content.
pnpm 9.8, Vuetify 3.7, Neo.mjs 7.0
Join Us for ViteConf on October 3rd — Learn how the best teams are building the next generation of the web with Vite!
StackBlitz sponsor
📒 Articles & Tutorials
50 TypeScript F–k Ups Mistakes — An admittedly colorfully-titled book digging into lots of subtle mistakes you might run into with TypeScript. It’s available on Leanpub in PDF, iPad, and Kindle forms, or you can read it all directly on its GitHub repo. At least worth a skim in case you’re running into any of its points..
Azat Mardan
The Official Redux Essentials Tutorial, Redux — The long standing guide to how to use the popular Redux state container the right way with best practices has undergone a big reworking with TypeScript used throughout, new concepts added, and more coverage of RTK/React Toolkit features.
Redux Team
React is (Becoming) a Full-Stack Framework — Is React merely a frontend library? How does the backend fit in? The author shares his thoughts on what led him to start considering React as more of a full-stack solution.
Robin Wieruch
📄 Using JavaScript Generators to Visualize Algorithms Alexander G. Covic
📄 Optimizing SPA Load Times with Async Chunks Preloading Matteo Mazzarolo
📄 Using isolatedModules in Angular 18.2 Thompson and Lyding (Angular Team)
📄 How to Generate a PDF in a JavaScript App Colby Fayock
🛠 Code & Tools
Milkdown: Plugin-Driven WYSIWYG Markdown Editor Framework — A lightweight WYSIWYG Markdown editor based around a plugin system that enables a significant level of customization. It’s neat to see the docs are rendered by the editor itself. GitHub repo.
Mirone
Fuite 5.0: A Tool for Finding Memory Leaks in Web Apps — A CLI tool that you can point at a URL to analyze for memory leaks. Here’s how it works. There’s also a video tutorial.
Nolan Lawson
✂️ Cut Your QA Cycles Down to Minutes with Automated Testing — Are slow test cycles limiting your dev teams’ release velocity? QA Wolf provides high-volume, high-speed test coverage for web and mobile apps — reducing your test cycles to minutes. Learn more.
QA Wolf sponsor
LogTape: Simple Logging Library with Zero Dependencies — I’m digging this new style of library that promises support across all the main runtimes (Node, Deno, Bun) as well as edge functions and the browser devtools.
Hong Minhee
📊 Chart.js 4.4: Canvas-Based Charts for the Web — One of those libraries that feels like it’s been around forever but still looks fresh and gets good updates. Bar, line, area, bubble, pie, donut, scatter, and radar charts are all a piece of cake to render. Samples and GitHub repo.
Chart.js Contributors
Legend State: A Tiny, Fast and Modern React State System — A year ago, Jack Herrington wondered if Legend State could be ▶️ ‘the ultimate state manager’ and things have progressed a lot since, with it now boasting being the fastest React state library in town.
Jay Meistrich
Tagger: Zero Dependency, Vanilla JavaScript Tagging Library — You can play with a live demo here.
Jakub T. Jankiewicz
tinykeys 3.0: A Keybindings Library in ~650 Bytes — Keeps things as simple and sweet as possible.
Jamie Kyle
heic-to: Convert HEIC/HEIF Images to JPEG or PNG in the Browser
Hopper Gee
Cheerio 1.0 – HTML/XML manipulation library for Node.
🎨 Chroma.js 3.0 – JavaScript color manipulation library.
eta (η) 3.5 – Embedded JS template engine for Node, Deno, and browsers.
Embla Carousel 8.2 – Carousel library with fluid motion and good swipe precision.
d3-graphviz 5.6 – Graphviz DOT rendering and animated transitions.
Alpine AJAX 0.9 – Alpine.js plugin for building server-powered frontends.
Happy DOM 15.0 – JS implementation of a web browser sans UI.
Elliptic 6.5.7 – Elliptic curve cryptography in plain JS.
Poku 2.5 – Cross-platform JavaScript test runner.
💚 Use Node? Check out the latest issue of Node Weekly, our sibling email about all things relating to Node.js — from tutorials and screencasts to news and releases. We do include some Node related items here in JavaScript Weekly, but we save most of it for there.
#702 — August 29, 2024
JavaScript Weekly
Rspack 1.0: The Rust-Powered JavaScript Bundler — Far from being ‘yet another bundler’ with its own approach and terminology to learn, Rspack prides itself on being webpack API and ecosystem compatible, while offering many times the performance. The team now considers it production ready and encourages you to try your webpack-based projects on it.
Rspack Contributors
💡 Rspack also has a family of ancillary tools worth checking out, such as Rsdoctor, a tool for analyzing and visualizing your build process (for both Rspack and webpack!)
Front-End System Design — Learn to create scalable, efficient user interfaces in this extensive video course by Evgennii Ray. Explore the box model, browser rendering, DOM manipulation, state management, performance and much more.
Frontend Masters sponsor
How to Create an NPM Package in 2024 — Sounds simple, but there are a lot of steps involved if you want to follow best practices, introduce useful tools, and get things just right. Matt Pocock walks through the process here, and there’s a 14-minute screencast too, if you’d prefer to watch along.
Matt Pocock
IN BRIEF:
🤖 v0 is an AI-powered tool from Vercel for, originally, generating shadcn/ui-powered React components based upon prompts you supply. Now, however, it has basic Vue.js support too.
Deno 1.46 has been released and promises to be the final 1.x release before the much awaited Deno 2.0. Deno’s Node compatibility improves even more (it now supports Playwright and many more things) and ships with V8 12.9.
📊 IEEE has published its latest annual list of top programming languages. JavaScript takes third place, but TypeScript has leapt up several places to fourth.
RELEASES:
Prisma 5.19 – The popular ORM for Node.js and TypeScript adds ‘TypedSQL’, a way to write raw SQL queries in a type-safe way.
📈 billboard.js 3.13 – Popular D3 chart library adds area-step-range charts.
pnpm 9.9 – Fast, space efficient package manager.
React Email 3.0, Ember 5.11, Bun v1.1.26
📒 Articles & Tutorials
JS Dates are About to Be Fixed — Handling dates and times is famously a painful area for programmers and JavaScript hasn’t done a lot to make it easier. Libraries like Moment.js help a lot, but Iago looks at how the Temporal proposal and its features will begin to help a lot more over time.
Iago Lastra
Weekly Chats on the Art and Practice of Programming — Your home for weekly conversations with fascinating guests about how technology is made and where it’s headed.
The Stack Overflow Podcast sponsor
JavaScript Generators Explained — Jan was frustrated by the quality of documentation and articles explaining generators in JavaScript, and set out to explain things in a way that a more advanced developer could appreciate.
Jan Hesters
Implementing a React-a-Like from Scratch — While it’s unlikely you’ll actually want to do this, at least thinking about it can prove instructive as to what’s going on in React’s engine room.
Robby Pruzan
▶ How to Implement the 2048 Game in JavaScript — Ania is back with one of her usual easy to follow walkthroughs of implementing a complete game in JavaScript. This time it’s the 2048 sliding puzzle game. (Two weeks ago she did Tic-Tac-Toe as well.)
Ania Kubów
Learn Role-Based Access Control and Simplify Permissions Management — Enhance security and streamline access by managing user roles with Clerk Organizations.
Clerk sponsor
📄 The Only Widely Recognized JS Feature Ever Deprecated – Spoiler: It’s with. Trevor Lasn
📄 Generating Unique Random Numbers in JavaScript Using Sets Amejimaobari
📺 21 Talks from the Chain React 2024 Conference – A React Native event. YouTube
📄 Exposing Internal Methods on Vue Custom Elements Jaime Jones
📄 The Interface Segregation Principle in React Alex Kondov
🛠 Code & Tools
TypeScript 5.6 Release Candidate — As always, Daniel presents an epic roundup of what’s new. We’ll focus more on it next week though, as the final release is anticipated to land next Tuesday (September 3).
Daniel Rosenwasser (Microsoft)
Vuestic UI 1.10: A Vue.js 3.0 UI Framework — Features 60 customizable and responsive components and with the v1.10 release it’s gained a significant bundle size optimization, a custom compiler that improves build time performance, and other minor enhancements. GitHub repo.
Vuestic UI
✅ Bye Bye Bugs — Get 80% automated E2E test coverage for mobile and web apps in under 4 months with QA Wolf. With QA cycles complete in minutes (not days), bugs don’t stand a chance. Schedule a demo.
QA Wolf sponsor
Material UI v6: The Popular React UI Design/Component System — At ten years old, the popular design system has its latest major release. There’s a focus on improved theming, color scheme management, container queries, and React 19 support. There are revamped templates to be inspired by, too.
García, Bittu, Andai, et al.
npm-check-updates 17.0: Update package.json Dependencies to Latest Versions — That is, as opposed to the specified versions. It includes a handy -i interactive mode so you can look at potential upgrades and then opt in to them one by one.
Raine Revere
Code Hike 1.0: Turn Markdown into Rich Interactive Experiences — Aimed at use cases like code walkthroughs and interactive docs, Code Hike bridges the gap between Markdown and React when creating technical content that takes full advantage of the modern web.
Rodrigo Pombo
Calendar.js: A Calendar Control with Drag and Drop — A responsive calendar with no dependencies, full drag and drop support (even between calendars), and many ways to manage events with recurring events, exporting, holidays, and more.
William Troup
📊 Perspective 3.0 – Data visualization and analytics component. The core is written in C++ and compiled to WebAssembly where it can be used from JavaScript. Their homepage shows it off well with a live example.
json-viewer 3.5 – Display JSON data in a readable, user-friendly way.
♟️ Stockfish.js 16.1 – A JavaScript chess engine.
jest-dom 6.5 – Jest matchers to test DOM state.
Marked 14.1 – Fast Markdown compiler / parser.
Javet 3.1.5 – Java + V8. Embed JS into Java.
Pixi.js 8.3.4 – Fast 2D on WebGL engine.
This September’s collection is packed with cool JS libraries, an experimental CSS framework, and other awesome tools!
Our resources collection for the month of October includes helpful fronted libraries, a cute CSS framework, and some awesome Node.js tools.
An amazing payment form widget, a new Vue.js framework for static websites, and lots more in our compilation for the month of November.
We’re kicking off 2020 with a list of some of our favorite web dev libraries, frameworks and tools that you should use in your next project.
Our web dev resources compilation for February comes packed with awesome React frameworks, tools for writing better tests, and much more!
New JavaScript tools from the Facebook developers team, an excellent Electron starter kit. and more in our web dev resources compilation for March!
A collection of our favorite web dev resources for the month of May, featuring useful JS tools, React components, and more!
have an annoyance that I i'd like to know a workaround for please:
-often/frequently/ nearly everytime i use MS 365 sign in or gmail sign in it takes me to a password reset. i have no wish to change my password { also previously did but it still kept wanting me to do so immediately after!]
how can i stop this please.
[working on a HP Z Book Win 10]
*sorry if this is a repeat question somewhere, i did a search and nothing came up
In today’s digital landscape, businesses are increasingly relying on a myriad of software systems and tools to manage their operations. However, as the number of these systems grows, so do the complexities and inefficiencies associated with them. This is where system integration comes into play. By streamlining your business through effective system integration, you can …
Continue reading "Streamlining Your Business: The Power of System Integration"
The post Streamlining Your Business: The Power of System Integration first appeared on Lucid Softech.
In the fast-paced world of startups, building a Minimum Viable Product (MVP) can be the key to validating your business idea, gaining early traction, and securing funding. An MVP is essentially the most stripped-down version of your product that still delivers value to your target audience. But creating an effective MVP requires more than just …
Continue reading "Essential Tips for Crafting an MVP: What Startups Need to Know"
The post Essential Tips for Crafting an MVP: What Startups Need to Know first appeared on Lucid Softech.
As a one-woman solopreneur running her own business, I’m constantly looking for new AI tools to help save time, increase productivity, and bring about a better ROI. The ecommerce sector is no different.
That’s why I’ve spent some time reviewing the best AI tools for ecommerce so you can get a leg up on your competition.
If I know anything about trying out new tools, it can quickly become costly. So, instead of spending precious dollars from your budget, bookmark our list of AI tools to use whenever you need to add something new to the stack.
Before we discuss which tools are the best, let's review why AI tools are helpful for ecommerce.
Table of Contents
I like to think of AI tools as “software on steroids.”
Ecommerce software helps streamline your business, but you can do so much more in less time when AI is added to the mix.
These kinds of tools are high-performing, meaning you can achieve results, gather new insights, or finish a task much more efficiently than if you were to do it by hand.
There are plenty of reasons to consider adding AI tools for ecommerce to your sales platforms. Let’s look at a few more benefits of adding AI to your tech stack.
In my opinion, customization and personalization are the way to go if you want to make a sale or keep a customer.
As technology advances and AI becomes more commonplace, your customers expect a better, more personalized experience. In fact, 73% of customers already expect it.
AI tools can help you create more personalized shopping experiences.
By analyzing your customer's data, such as interactions and behavior, AI can help you tailor product recommendations and content and create more targeted marketing campaigns.
The more personalized the recommendations or marketing campaigns are, the more likely you’ll gain a conversion or a sale.
It’s no secret that chatbots are a staple for any ecommerce business. Chatbots can help answer questions or suggest products and services to your customers.
While chatbots are great, regular, run-of-the-mill chatbots come with limitations — specifically, canned responses.
AI-powered chatbots, though, are intuitive.
Because they’re trained on your customers’ data and constantly monitor new data as it comes in, they can anticipate customer needs and questions and continually improve responses with each interaction.
Another benefit I’ve found is that they can also help collect and sort important data, like contact information and interests, so you and your team can easily create customized mailing lists.
(Remember the importance of personalization? This can help!)
Your marketing strategy is only as good as its ROI. And if you’re not using your customer data and insights to create a solid marketing strategy, what are you doing?
Of marketing leaders who have already implemented AI into their marketing campaigns, 63% say their ROI has increased since investing in AI-enabled software.
So, what makes AI tools better than your current non-AI marketing tech and better for your ROI?
Analysis and predictions.
AI tools can help you uncover insights you might miss in your customer data.
Using those valuable insights, AI can help you predict and forecast customer behavior and marketing trends, helping you and your team create a better marketing strategy with a higher likelihood of success.
Depending on your ecommerce business, you can likely pinpoint when certain services or products are going to be a hit among your customers. For example, holiday-focused products will likely become more popular during the “Ber” months (you know, September, October, November, and December) than compared to the Spring and Summer when the interest is just not there.
But what if something happens, like a news event or an influencer push, that causes your product or service to become popular overnight, and you’re caught with limited inventory? It would be nice to be able to predict demand, right?
AI tools can help you make better predictions and forecasts using historical and real-time data. The insights your AI software provides can help you make better decisions on increasing or decreasing your stock and inventory, depending on customer behavior. AI can also help you anticipate supply chain operations so you can make a plan to improve logistics and demand.
One of AI tools' biggest benefits is the time they can save your team.
Of marketers who use AI tools, 86% say they save over one hour a day by streamlining their creative tasks.
Customer service teams who use AI chatbots say they help them save an average of two hours and twenty minutes a day.
I also suggest checking out how generative AI can help your content creation efforts.
The time saved using AI software can be better used elsewhere, like interacting with customers and creating a personal relationship with your brand.
While AI tools can save your team time and help you create better, more personalized interactions with your customers, I feel I must tell you there are a few limitations you should be aware of, like:
Although these limitations might be a con, the pros of AI tools for ecommerce far outweigh them, in my opinion. Let’s take a look at some of the best AI tools for ecommerce on the market today.
Content marketing is a key element of your ecommerce strategy. A continually updated blog can help drive sales, educate your audience about your products and services, and help you rank in the SERPs.
Content creation can take significant time and effort, so I suggest adding Content Hub to your ecommerce stack.
Content Hub is your all-in-one AI content marketing software. With Content Hub, you can quickly and efficiently generate content ideas, launch A/B testing, easily include lead capture tools in your content, and gain insights into user behavior.
Features
Pros
Cons
What I like: If you’re already using the HubSpot CRM, using Content Hub for marketing- and content-related purposes makes sense. I like that all aspects of your ecommerce business, from sales to content creation to customer outreach, are housed in one customer-focused CRM. Say goodbye to multiple platforms!
Pricing: You can try Content Hub for free. If you like Content Hub and want to access more features, Starter plans start at $20 a month per seat.
What if there was an AI tool that could help you identify holes in your customer journey, better understand your customers' mindsets, and suggest new products and services to them?
The good news is that there is such a product, Octane AI.
Octane AI is an ecommerce tool that helps you get to know and better understand your customers. From collecting their contact details to getting a feel of their preferences, desires, and concerns, Octane AI can help you plan a better, more personalized customer experience.
Plus, I found out that if you operate a Shopify store, you can incorporate this tool into your site to quiz customers on their preferences, and the results will give them product recommendations based on what they like.
Features
Pros
Cons
What I like: Octane AI can help improve your personalization efforts for marketing campaigns. I like that each quiz can help segment customers into various groups, so I know which groups care about which products. Using that information, I can easily send targeted messages to various segments.
Pricing: Octane AI starts at $50 a month for 200 engagements. Octane Plus, which offers more features and engagements, starts at $250 a month.
SearchIQ is AI software that helps deliver better customer search results.
With this tool, you can elevate the customer experience and increase user engagement while gaining valuable insight into the kinds of products and services your users are searching for on your website.
I think this tool is best used by ecommerce businesses with many product offerings and website pages.
If you offer hundreds of thousands or even millions of products, you can still use this tool — impressive, I know. It can easily handle and sort through all of your products and services.
Features
Pros
Cons
What I like: Sometimes, I am the most impatient person on the planet. When searching for a product, I don’t want to spend an unreasonable amount of time searching for it. I like that SearchIQ offers lightning-fast suggestions, which helps those impatient customers like me stay on my website longer.
Pricing: You can try SearchIQ for free. Plans with access to more search features and reports start at $19 a month.
Customer retention is a critical factor in any ecommerce business’s success.
RetentionX is an AI tool for ecommerce businesses that can prevent churn, increase retention rates, optimize customer acquisition channels, and automate certain workflows.
RetentionX helps you easily collect and analyze valuable customer data. This data can be used to create an action plan that best supports your customers and ultimately transforms them into loyal and dedicated fans of your brand.
Features
Pros
Cons
What I like: RetentionX is a great tool for segmenting customers based on user activity. I appreciate this tool's detailed overview, which helps me better understand which percentage of my customers interact with my brand and which needs extra care. This information can be helpful for planning my marketing strategy.
Pricing: If you have less than 250 active customers, you can try RetentionX for free. After that, you’ll pay $0.01 per customer.
Several parts of a marketing strategy can be automated, and often, those parts offer the best insights and data.
Marketing Hub, a HubSpot ecommerce software, is AI-powered marketing software that allows you to automate your strategy and generate high-quality leads efficiently.
Integrating Marketing Hub into your CRM means all of your data and tools are in one place. Use this tool to help drive revenue, save yourself time, and measure and optimize your customer journey and marketing campaigns.
Features
Pros
Cons
What I like: I like the ease and convenience of having my sales and marketing tools all in one place. Marketing Hub easily combines with the sales platform to help me better understand my customers and save me time with the automation features. This is a big pro for any sized business, whether big or small.
Pricing: You can try Marketing Hub for free. Try the Premium Starter version for $20 a month per seat for more features.
Running an ecommerce business involves many moving parts, and it requires teamwork and collaboration from every member of your team to ensure its success. ClickUp is a one-stop shop for all things collaboration, from project management to product development to marketing and sales.
I appreciate that its AI integration allows users to summarize meeting notes and threads easily, so your team can stay on top of the nitty-gritty details of projects and product launches.
Features
Pros
Cons
What I like: As they say, “Teamwork makes the dream work.” I appreciate that ClickUp’s focus is on providing a collaborative workspace for team members to collaborate to provide the best customer experience. The AI summarization is also nice, because it helps to create an easily skimmable summary of threads and meeting notes — which can help save time and effort!
Pricing: You can try ClickUp for free. For small teams, it's $7 a month, and for larger teams, it’s $12 a month.
Good copywriting is a must for any ecommerce business.
You’ll need product descriptions, landing pages, and blog content if content marketing is part of your marketing strategy. It can take some time to write all those pages. Surfer AI is here to change that.
Surfer AI helps business owners create optimized and researched content in minutes. It can also be used for competitor analysis, keyword research, and content auditing.
Features
Pros
Cons
What I like: AI-written content is easy to detect. As great as ChatGPT and other content writing services are, the overuse of words like “delve,” “boost,” and “transform” is almost a dead giveaway that the content was created using an AI tool. I appreciate that Surfer AI includes an AI detection and humanization feature. This can help your content feel less robotic and more personal.
Pricing: Surfer AI is $89 a month for small business owners and freelancers. A subscription is $219 a month for those looking to scale their content with AI.
Pricing is another important aspect of your ecommerce business.
You don’t want to overcharge your customers, nor do you want to underprice your products and services. What if there was a way to monitor competitors' prices so you can confidently keep your prices competitive?
Enter Prisync.
Prisync is a competitor price tracking and monitoring software. Gone are the days of spending hours tracking down competitor pricing by hand.
Instead, with Prisync you can see all of your competitors’ pricing and their stock availability all in one interface.
Features
Pros
Cons
What I like: Comparing competitor pricing can be tedious. And depending on your market and product offering, prices can change in an instant. I appreciate that Prisync keeps a tab on changing prices in real time. This way, it’s easy for you to lower or raise prices depending on the current supply and demand.
Pricing: You can monitor up to 100 products for $99 a month. If you wish to monitor more than 100 products, sign up for the Premium plan, which starts at $199 a month.
An AI chatbot can act as a virtual assistant to help your customers and website visitors.
With HubSpot’s AI Chatbot Builder, you can quickly and easily build an intuitive chatbot to monitor and review leads, book meetings with prospective customers, and provide helpful customer support.
Plus, if your chatbot is unable to answer a customer’s question, it can easily connect your customer with a live team member to further provide assistance.
Features
Pros
Cons
What I like: I am a big fan of AI, but sometimes its capabilities can be limited. I appreciate that HubSpot’s AI Chatbot Builder allows for the option to provide live chat support. This can help create better, more personal connections with your customers while still keeping your agents free from other tasks that can be automated.
Pricing: You can try the chatbot builder for free!
I know from my own small business that growing your ecommerce business is an exciting opportunity, but it often comes with a lot of work to make it a success.
AI tools for ecommerce can help automate certain tasks to free your team from more important items on their to-do list, like personally interacting with customers and solving their complex issues.
Ecommerce AI tools aren’t just for customer support, though. You can use them to optimize your landing pages, create SEO-ready content, or monitor competitor pricing. If you’re ready to grow your business, check out HubSpot AI-enabled ecommerce tools today.
