Driving Business Results with Appsmith AI and New AI Integrations

 Bharath Natarajan
Posted by Bharath NatarajanPublished on Feb 14, 2024
9 min read
Announcing Appsmith AI Blog

The release of ChatGPT and its seemingly endless ability brought AI into the mainstream. Everyone from news organizations to thought leaders and industry figures was speculating (often wildly) about how AI would forever alter how businesses operate and interact with their customers. After all, there’s a lot of money to be made in selling AI tools with big claims and little substance. But how do things look now that the dust has settled and reality has set in?

The answer: AI’s tangible business benefits have been clearly demonstrated, especially when it comes to streamlining business processes, applications, and customer interactions. This article explores the business use cases we’ve seen impacted the most by AI and how Appsmith AI and the new AI integrations makes this achievable for organizations of any size.

AI isn’t a miracle solution to all problems, but it can save your business time and money

We’ve seen a lot of promising use cases over the last few months as developers have started building their own AI-powered applications with Appsmith AI and our external APIs like OpenAI, Google AI, and Anthropic. One major example is an AI Google Meets assistant that automatically captures notes and action items from meetings  — demonstrating how AI can be used to save employees time and improve efficiency.

Our main takeaway from this and other real-world applications of generative AI is that it is a powerful business tool. It won’t run your business for you, make decisions, or design your next product, but it will save you time and money if used effectively. For this reason, we believe that it’s going to become an integral part of many business workflows.

That’s why we’ve added support for AI natively in Appsmith. This allows users to easily integrate AI into their applications to streamline their workflows, improve decision making, and gain insights from data more quickly with AI-driven apps built on Appsmith.

The top real-world AI business use cases

AI in its current form, mostly powered by large language models (LLMs), isn’t perfect; but with the right techniques and processes, these tools can save huge amounts of time (and therefore money) as team members no longer have to do many lower-leverage tasks and can instead focus on higher-leverage ones.

Here are some of the AI applications that we’ve seen our users build in Appsmith that are making a real, tangible difference to their business operations.

Generating emails

It’s no big secret that a lot of the emails we send daily, both internally and when communicating with customers, are largely the same, with small details differing between them. AI is perfect for helping craft these emails. It can be supplied with the details of a customer query, order, or other information, and fill in the text around it to turn it into a polite, helpful message.

What does this mean for business? Faster response times, more satisfied customers, and less employee load. Customer support and sales staff can be presented with an AI-drafted email and only need to confirm or tweak the details before sending it, rather than writing the whole email themselves. If the email is inaccurate or some nuance is required, they can override the response — and have the AI learn from their changes for future drafts.

Screen recording showing AI’s ability to generate text with specific tones and styles.

Generating support ticket responses to help customers

This is already a widespread use case for AI text generation. Imagine an AI customer support tool that can read through a customer's communication history (including text messages, onsite chats, and even phone calls) along with any documents linked to support tickets (such as a PDF showing their order that never arrived), allowing them to automatically suggest accurate, relevant responses to customer questions. As with email generation, even if these suggestions need to be tweaked, there is still a major increase in productivity through templating responses.

This use case demonstrates the benefits of text extraction performed by AI — letting it retrieve data from unstructured data sources. Another example of this is extracting items and costs from reimbursement receipts and checking if they match the company’s return policy, allowing further automation of the customer experience. There are many aspects of customer support like this that could be either partially or completely automated.

A similar use case would be allowing AI to automatically read through a project’s entire codebase and create a first draft of project-wide documentation based on a predetermined format. This documentation could be automatically updated in every pull request and easily reviewed just like normal code changes.

Creating product descriptions and ad copy that drives engagement

This is yet another use case for AI’s text generation abilities. AI can be helpful for templating, as in the previous two use cases, and is also beneficial for brainstorming product descriptions and other kinds of text.

Imagine an AI tool that automatically reads through the customer reviews for your product and makes suggestions about how to craft product descriptions and ad copy to better match your customers’ own language to drive more sales. You can provide broad messaging and product details to the AI, ask it to present the copy in different styles, and bounce ideas back and forth with the AI to better craft your message for your product description or other sales material.

Streamlining customer communication with summarization

AI can be used to automatically read every support ticket or online review from your product, categorize the sentiment (happy customer or angry customer), and label the component of the product or service that they did or did not enjoy.

Screen recording showing a business AI application that reads through customer reviews, performs sentiment analysis, and classifies the department related to the review.

With AI, team members would no longer have to sift through a huge mountain of customer reviews or support tickets to find relevant information. The AI could instead look through the haystack in a few seconds and tell you where to find the needles based on your prompts. By constantly feeding your customer feedback to an AI platform, you could also receive instant notifications when a negative review pops up, so that you can quickly correct issues, maintain customer relationships, and constantly improve your product.

A person needs to dig through a large pile of reviews to find important ones. A robot, representing AI, can find the important information almost instantly.

It’s much faster when AI tells you exactly where to look to find the customer information you need.

This type of data extraction could also be used to improve coordination across teams to reduce siloing within an organization. Consider a tech company that is continually adding new features to their product and wants to keep the customers on their email list informed about them. Instead of the product marketer having to read through a mountain of JIRA tickets and GitHub PRs to create their email release notes, AI could automatically read through them and generate a first draft for them.

Extracting (and autofilling) information from uploaded documents

AI doesn’t just handle text — it’s already effective at creating and modifying images (and strides are being made on the audio and video front, too). Image extraction is a key use case for AI, as it allows converting information in an image to a string, which can then be further processed with conventional methods. This is something that seems simple, but has only recently become feasible with AI image models.

The loan processing application below showcases this functionality by allowing users to upload images of personal identification documents. Then, instead of requiring users to fill in the same data manually in fields of the application, AI can extract information like name, date of birth, and age from the image and autofill it. AI could also be used for similar applications that require uploading images of serial numbers or bar codes in IT asset or warehouse management systems.

Screen recording showing a business AI application that extracts a customer’s information from an uploaded image of a personal identification document and autofills the fields of a loan application.

Generating images (and saving on studio costs)

Those lacking in artistic skill can now quickly draft and iterate visual ideas using generative AI before sending off their drafts to graphic designers to create the final image. This drastically decreases back-and-forth and miscommunication throughout the process, allowing anyone to realize their creative ideas in a way that others can see and work with.

If you are a fashion designer, think about the convenience of creating images of models wearing your clothes without conducting a photoshoot. Compare the many-week-long process of calling models, scheduling photoshoots, hiring photographers, etc. to the alternative. Generating an AI model wearing your clothes could take only a few minutes, be much cheaper, and be almost (if not completely) indistinguishable from a real photo.

Appsmith lets you easily integrate AI with your business applications

These are all real use cases that have been built on Appsmith, not hypothetical scenarios and hype. Appsmith is already being used to streamline customer support processes. Now, whether you’re building new bespoke tools or extending your existing ones, you can enhance them with generative AI functionality using Appsmith.

As a low-code application platform, Appsmith allows you to interact with data from separate tools through integrations (like REST or GraphQL APIs), process them, and send the outputs to other tools through different integrations.

After building your AI-powered applications, you can embed them into other applications and web pages. Unlike many other low-code platforms, Appsmith is open-source forever, so you can feel comfortable building your business infrastructure on top of it.

Appsmith’s AI-powered app studio

Several months back, we introduced the Appsmith coding assistant, which allows developers to create complex JavaScript expressions from natural language prompts. We’re continuing to expand these capabilities to generate entire functions for JS Objects or SQL queries as well so that even users without programming experience can create interactive applications on our platform.

We’ve also recently introduced Appsmith AI – a fully managed AI integration provided through the Appsmith platform itself. With Appsmith AI, you can handle all of the use cases we’ve discussed so far immediately without an API key to an external model provider. You can also easily upload your own PDF or Markdown documents to provide context to the AI models that you query in your application.

We aim to support as many proven AI tools as possible, as we want developers to always have access to the best AI tools for their applications. That’s why we’ve created integrations for the largest third-party LLM providers like OpenAI, Google AI, and Anthropic as well. You can access any of these integrations (as well as ones that we add in the future) as long as you have an API key from the model provider.

We’re dedicated to making AI accessible to developers who need to be able to rapidly build and prototype apps. We see AI becoming a necessity for business, rather than an enhancement, so we’re constantly working to make sure our app platform provides the tools needed to easily make AI a central part of any application.

AI is ready to improve your business productivity and efficiency

Companies of all shapes and sizes are jumping on the AI bandwagon, and it’s not just because of the hype. Organizations that choose their tools carefully are seeing great benefits from that decision, as AI has become integral to the improvement and efficiency of many of their business processes.

Successful businesses will continue to see AI become more embedded in their processes and applications, especially as the underlying technology becomes more reliable and efficient. Choosing an app platform that recognizes the importance of AI is critical if you want future-proof applications that can be easily extended as new AI tools are released.

Introducing Appsmith AI

You can see more about how you can get the most out of Appsmith’s AI features with our Appsmith AI step-by-step guide.

Beyond the hype - AI use cases that are delivering real business value

140 IT and development teams share how they’re integrating AI capabilities into their internal tools and business processes, and which ones are proving the highest returns.