How Low-Code/No-Code tools can offer many advantages

If you need to design and build systems that involve data extraction, translation, and loading (ETL), low-code/ no-code tools such as Safe Software’s FME platform is well suited to help you along your journey.

In a recent podcast with Enterprise Management 360, I spoke to John Santaferraro, a seasoned data management expert, about how low-code/no-code tooling offers a clear breadth of advantages, ranging from ease of use to rapid POC creation and iteration.

In this article, I will provide some wider context to the story but I can’t promise that I won’t be mentioning the Great British Bake Off.

Firstly, a little background, I recently joined Tensing, an International group of geospatial specialists. Tensing has grown rapidly in recent years to provide experienced GIS and data specialists to help governments, utilities, and energy companies to make their data more valuable. We are now growing in size in the UK, France, and Belgium and collectively have the highest number of FME Certified Professionals in Europe.

Before this new role, I worked in the Oil and Gas service industry for over 15 years supporting the delivery of geoscience datasets to upstream oil and gas clients.

Low code, no code for data pipelines with all kinds of data? Is it real or is it just a buzzword?

Low-code/no-code is nothing new, ever since Excel, we could manipulate tabular data through macros or embedded VBA snippets. Software developers like Safe are just building on this. Microsoft has recently embraced this with tools such as PowerBI, Power App, and Automate. Users who are willing to take the time to think if there is a smarter way/or little hack to get a task done faster are now coined by Microsoft under the term 'Citizen developers' - I just call them ‘power users’.

One of the key reasons I enjoy using the FME Platform for building spatial integration solutions and ETL tasks is that it is built on a low-code/ no-code approach coupled with a visual building block approach. Each functioning component or ‘transformer’ in an FME project is effectively a visual programming building block that pieces a data transformation together. I am a very visual person, so this type of development environment works for me and indeed it is no surprise this software developed from and has a stronghold in spatial data ETL.

The Great British Bake Off analogy…?

It is at this stage I bring in the baking analogy. As a fan of the Great British Bake Off watching rather than emulating, I couldn’t help but see some similarities between coded solutions and low-code/ no-code tools.

Anyone who has watched the show is going to be familiar that each week there are three challenges, one of which is the technical challenge. The bakers are challenged to replicate a textbook-perfect cake with no warning of what they will be baking. They are given the ingredients, the tools to bake with, and of course a kitchen to work from.

In many ways, these are like the data challenges users face all the time. If you swap out the ingredients for data sources, the textbook cake for the end data output, and the baking tools with your code or software. Maybe you can see where I am going…

The equivalent of coding a solution from scratch would be to imagine having the contestant bakers not only had to produce the perfect cake but beforehand also needed to unpack all the kitchen utensils, dig out the best-sized cake tray and wire up the KitchenAid mixer - all before starting to mix the ingredients and get baking.

Using a no-code/low-code solution, you can imagine the kitchen to be all set up and ready to go, giving plenty of time for the bakers to focus on the real task at hand and all the while having access to a good recipe book.

Why are data teams still doing so much coding? What are the potential pitfalls?

There is no doubt that there are some fantastic Python toolkits and solutions. If you have a talented team around you with the time and resources to develop and maintain there are very good reasons for developing fully custom-coded solutions. Unfortunately, what is often the case is project-based tasks that come and go, have limited completion times with set end dates, and have a lack of resources to provide ongoing support.

With the recent ‘Great Resignation’ there is a very active job market at the moment, so to take it back to the baking analogy.

If you were to lose your skilled French pastry chef from your bakery you are in trouble. A new chef may not be willing or skilled to prepare the same speciality cakes. Indeed, if you lose a key programmer in your team, others probably won't be able to pick up immediately debugging the code or indeed won’t be willing to work with someone else’s coding style.

Moreover, if you don’t look after your kitchen tools they will fail over time. This same can be said of dependencies on code libraries - you need to update and maintain them or risk security threats and compatibility issues.

Safe Software maintains and develops its FME platform, and in effect takes over that role of ensuring that a solution built today will continue to be forwards compatible with its platform years into the future. They continue to build data readers and writers for new data formats as they appear and maintain support for over 450 formats.

Why use FME as a low-code/no-code tool

In my previous job as a technical team lead, a key goal for me was to ensure that people with the domain knowledge had the tools and ability to undertake the data transformations that they require, rather than being dependent on other teams or resources. Effectively supporting the team to build their self-service data projects and solutions.

To that end, FME has a shallow learning curve and is supported by broad training materials, a very supportive community, and partners including Tensing who span the globe.

Over the years I created a diverse range of FME workbenches working with a variety of formats. For the most part, I was able to run through each workbench step-by-step at a reasonable pace. There were a few notable features of FME that helped me along the way:

• Feature caching is one of the best features of FME, enabling you to view data output through each building block transformer in the project. Having the ability to step through and view the data transform one step at a time means fewer whiteboard and hand-waving moments.
• Self-documenting capabilities - you are effectively building the data-wrangling recipe steps as you go. Annotations you can put text annotations in workspaces either associated with transformers or floating. It’s a great way to document key how’s and why’s directly within the workflow.
• Bookmarks going back to my school days I was always told to work through problems in bitesize chunks. Bookmarks are the means to break apart workspaces into clearly defined sections and move multiple transformers relating to the same task at once. Importantly and unlike most code they can be reused and repurposed.

These features of the FME workbench combine to make a self-documenting solution which meant during my handover period projects were easily handed over to colleagues.?

Why has Tensing chosen to be a Safe Software Partner?

A few years back Tensing decided to restart and decided to focus purely on consulting services rather than software product development. This enabled the team to work with the most advanced technology partners and build a team of passionate spatial experts.

At Tensing, the focus remains on driving geospatial innovation, the key to this is to help to democratise the use of ETL and GIS technology. The FME platform with its visual low-code/no-code approach helps to remove many of the barriers to reaching this goal and supporting our customers in building up spatial ETL and GIS skills.

Who is best positioned to benefit from no code/low code tools and solutions?

If you are undertaking repetitive and low-level tasks involving data wrangling from a range of spatial data sources there will be quick wins that Tensing can help you to attain with FME and other tools. By supporting you in FME projects that free you and your colleagues to focus on more taxing, higher-value challenges. This also helps keep team members engaged, collaborative and looking forward to working on new data challenges and enables more of the team to support and contribute resulting in a lower chance of resource bottlenecks.

No-code/low-code solutions are also useful in the development of quick turnaround proof of concept and minimum viable product solutions, using FME Server you can create FME Apps click on the image above and open an app or two too discover what is possible. Anyone working in such an environment who is encouraged to move fast and break things will enjoy the agility that comes with working with the FME platform.?

If any article strikes a chord with your current data challenges, feel free to get in touch.

A special thank you to Alessia N. for her illustration used in this article.

#fmepartners #lowcode #nocode #fme #elevatingspatialdata