The New Front-End Landscape

This is post 4 of 4 in the series “The New Front-End Landscape”

Data is (and is not) Like Oil
Are You a Data-Intensive Application? [Yes]
Building Non-Linear Applications
The New Front-End Landscape

Data-intensive applications are everywhere. Companies are building these products to sell, and increasingly, more companies are building them internally to leverage their own internal data, AI/ML initiatives, sensors, and more.

We have seen countless companies sitting on exabytes of data but not having the ability to turn oil into gold. You constantly hear about companies that are never able to leverage their customer data, create a new offering for their customers, or even internal departments being stymied by their internal IT’s resources and bandwidth to take on new projects. Often this is because the bottleneck for creating such rich and non-linear workflows has been the creation of a usable front-end.

When we look out at the front-end development landscape we can decompose the world into 2 axes… how fast a team can get value/build a product and how open is the stack.

And in this space, something very exciting is happening at Kleeen Software – at the intersection of an open-stack and fast time to value.

Manual

Manual Development:
Open Stack + Long Time to Value

This is the current way companies build software today: Humans – lots of talented humans. Human designers make mockups and wireframes. Human developers convert those designs into code. This is a process that takes months to execute. And it isn’t just creating the MVP; every new feature and function you build must be designed, maintained, updated, and enhanced. And while some companies or IT teams forgo design, that can actually cost more in the long (and short) run. Even with component libraries, visualization libraries and advances in JavaScript frameworks, manual UI development is a long, erratic process, often focusing on front-end infrastructure and common workflows rather than on domain or product-specific features that can “wow” and win-over your customers.

But while the time-to-value is long, the control is quite high. You and your team can customize everything. The front-end stack is yours to control… and so is your middleware, your backend, your DB, your hosting, authentication and more. Anything you ask your engineering team to do is technically “possible,” but it all comes at a cost. That is the freedom that has kept manual development the go-to solution for building data-intensive SaaS applications.

Low-Code:
Closed Stack + Fast Time to Value

Part of the Low-Code “revolution” over the past few years has been the adoption of WYSIWYG tools – where product creators can drag and drop UI elements into pages. However, if you do not know the best practices to create good UI/UX, these tools quickly let you build bad UI/UX. Moreover, every link, every action, every table must be manually created – and the creator using these tools frequently ends up having to write code almost immediately to connect data needs to the backend provided.

And speaking of back-ends, most of these solutions are black-box monoliths – locking you into their stack. You are now limited to the database, data structure and infrastructure of those platforms. Your data must be stored on their systems, exposing you to any security risk that the vendor has – and you do NOT get your front-end code. These front-ends run on their cloud.

Because these tools are so limited in their front-end flexibility, each of the platforms in this space tend to focus on a specific type of application – the majority are data-entry and form-based workflows.

“most low-code platforms are in their own development silos, separated from other development tools with their own IDEs…Even the biggest and best low-code vendors admit that just using low-code isn’t going to power your business” – TONY GRAHAM

All these limitations have created questions — can low-code really rise to the expectation and promise advertised? They have found a new niche, but that is what it is – a niche – not a solution to power your software development needs. In that niche, you can move very fast – but a low-code platform is a black-box solution.

Embedded BI Tools:
Closed Stack + Slow Time to Value

When BI and Dashboarding tools were first created, they were never intended to deliver a product to end customers – they were data visualization tools. And they were fantastic. Load your data in, do aggregation and “basic analytics”, and visualize your data as a dashboard or a report. Now users that were not DB experts could slice their data and garner insights – once your DB architect sets up your BI tool, loads your data into their proprietary system, and teaches you how to use the BI tool itself.

While powerful for the right use case, this isn’t usually the quickest of solutions to deploy. Also, end-users found that BI-tools have limited utility for empowering knowledge workers to take action. The visualization of data is there to provide context which is meant to help an end-user to reach conclusions on actions to take, but these knowledge workers must switch between multiple applications to get insights and take actions to get their job done. This swivel-chair approach slows down end-users and increases friction. As a result, no one was fully satisfied with BI tools, but they were the best solution available.

As more and more companies started building their own data-intensive solutions for customers or internally for their employees, BI tools wanted to capture a piece of the pie too and started advertising “Embedded BI” – the idea of embedding a BI tool’s visualization or dashboard in your application. But let’s think about that. Now you still have to build all your product’s infrastructure, you still need to build the pages, dashboards and support all the rich features you need to ship a product – but you can save time by using a BI tool to make a visualization. That can be useful, but there is still a lot of manual development work to create everything that is not those BI visualizations – you must expose all your data to the BI tool’s proprietary database/stack/cloud, and the embedded BI components are not truly integrated with your shipping app. You can’t cross-link and connect/style those elements to the rest of your product.

In short, embedded BI takes a huge amount of time, requires much manual work, adds unnecessary complexity to your architecture, and still locks you into a closed stack with data-duplication/privacy issues.

Kleeen Software:
Open Stack + Quick Time to Value

Kleeen Software has taken a wholly different approach. Our solution is not about enabling customers to build workflow wizards or forms for their end-users; rather it empowers customers to build products/solutions which the end-user can use to investigate data a little differently every time. This is perfect for unstructured and non-linearly structured tasks – where there exists a clearly defined structure and relationship to the underlying data.

To create a product with Kleeen, you input WHAT you want your customers to do and WHY – not drag and drop. And as a result, it isn’t just “low code,” Kleeen is No Code to generate the application front-end. Kleeen’s AI will construct workflows, design UI, and generate production code ready for you to integrate to your APIs and ship.

Moreover, instead of previewing with a mockup, picture, or drag-and-drop layout – Kleeen builds you an instant prototype – the real production UI code (running on Kleeen’s cloud) but with simulated data. Test and validate before you spend months building out a data-lake or analytics. This is true WYSIWYG because this is the real UI with all interactions there to experience live. Thus you can design and build out a new product in a few hours, iterate and have a full front-end integrated with your stack in days, not months.

Kleeen Software is not just about BI and data visualizations. We create the UI and the UX, allowing you to create a “Single Pane of Glass” to merge insights and context from those data visualizations with CRUD operations, actionable insights, and buttons that can connect to complex back-end business logic or even change data values to move a company forward. In short, we allow end-users to analyze data, make their decisions, and take action, regardless of where the data resides.

The secret here is transparency. Kleeen Software generates front-end code and checks it into your Git repository as a pull request. You can see, review and approve any and all coding changes. Nothing is closed. Because at the end of the day, your team is responsible for your product – which means you are responsible for the code. And Kleeen Software knows that too.

In addition, every API needed by the generated front-end has a corresponding hook providing state and context from the UI and providing you the specification for the incoming data from whatever backend technology you wish to hook up. In the end, Kleeen is just a developer on your team — an extremely highly productive one.

Furthermore, Kleeen allows you to drop in custom modals, widgets, pages, authentication systems, even styling. As a result, if you like your BI tool, drop in embedded-BI. If you need to create something custom that is domain specific, product specific, or a dashboard with “sizzle,” Kleeen doesn’t stop you at all. Want to restyle your product to conform with your brand, go ahead and drop in your custom CSS. Kleeen doesn’t box you in – it takes the most common workflows and front-end infrastructure off your plate with a true no-code solution, freeing up time, people and money to focus on what makes your product great. And next time you have a new feature, want to make an update, or respond to a customer – it turns into a resourcing decision not an “if” decision.

Cover Photo by Robert Lukeman on Unsplash