Low code application platforms (LCAP) – How to avoid costly mistakes while achieving rapid application delivery goals?
Vinay Mummigatti
C-suite Executive & AI Thought Leader | EVP at Skan.AI | AI, Automation & Digital Strategy | Enterprise Transformation | Driving Scalable Growth, Innovation & Operational Excellence
Low code application platforms (aka LCAP) are a powerful tool in the current technology landscape. The ease of learning, speed of development, and ongoing ease of maintainability make LCAPs a compelling alternative to bespoke coding or customizing complex enterprise applications for a large number of business applications. I have used multiple Low code platforms extensively and scaled them at the enterprise level. My experience of LCAP’s at large companies like United Healthcare, Bank of America, and now at LexisNexis has taught me to balance the advantages of LCAPs with the restrictions/limitations they bring. I am writing this article to share some of my experiences which may help the readers to avoid costly mistakes in their LCAP journey.
?A ?Low-code platform is characterized by its use of model-driven or visual development paradigms, declarative techniques, expression languages supported by scripting instead of traditional programming. The latest LCAP tools are omnichannel ready and offered as SaaS.??
?Many software vendors are jumping on the bandwagon of labeling their platforms as LCAP – much to the peril of misguided adoption of LCAPs for patterns that result in failed project outcomes. Some challenges I have seen with LCAPs’ are also due to the overzealous product vendors who position their LCAP products as a panacea for all kinds of solutions, which result in major problems with?scalability, performance, integrations, security, or usability.
Need for LCAPs:
A typical corporate technology landscape is a mix of many complex homegrown applications, legacy platforms, ?and complex third-party vendor platforms like ERP or CRM. As many of these systems are “built for purpose” with elaborate hard-coded logic, it is difficult to make changes rapidly or economically. There are many business requirements driven by regulatory compliance, operational efficiencies, quality, customers, or cycle time improvement, which require us to develop custom applications. These applications could be meant for orchestrating work across systems and departments with real-time visibility, aggregating data from diverse sources for meaningful actions, automating decisions and flows, collaborating or eliminating non-value-added manual work from hundreds of back-office tasks. To address such needs, the business teams want IT to deliver rapid solutions that are easy to model, configure, build, manage and deploy without lengthy change management cycles and IT dependency. LCAPs offer a compelling solution to such needs.
Emergence of Citizen developers :
?The tech-savvy business teams want to go one step further – they want to lead the application development and manage ongoing changes using minimalistic support from IT departments. A new breed of citizen developers is emerging - Citizen development is business users with little to no coding experience building simple business applications with IT-approved technology like LCAP to improve their day-to-day productivity or an existing business process. Citizen developers?are increasingly demanding LCAP based solutions that use model-driven declarative development.
Key features and value of LCAP :
?The biggest differentiators for LCAP tools as compared to other traditional tools are?:
?1.????Graphical modeling environment to define business logic, flows, and requirements
2.????Configurable processes, rules, roles, privileges/access controls, reports, and forms
3.????Real-time visibility to work status with out of box audit reporting
4.????Pre-built connectors / API’s with industry-standard platforms such as CRM, ERP, RPA, Middleware, Content management systems etc
5.????Omni-channel ready User experience
6.????Pre-built and easy to configure reporting and business activity monitoring
7.????SaaS ready
领英推荐
8.????Self-guided design, development, and testing tools with in-built best practice and guardrail checks in real-time.
9.????Abstraction of all application layers with composable architecture
10.??Dynamic change capability for business users to configure the functionality with limited IT engagement – enabling continuous improvement and optimization of business processes.
A couple of other significant advantages of LCAP are :
1.????Business-IT alignment (Beyond agile) : Most LCAP tools offer the ability to capture requirements from business users directly into the modeling environment which are then converted into executable design elements. Hence, the business and IT teams can collaboratively capture requirements, develop and demonstrate prototypes as an iterative process. ?This is an immense value for business users as they don’t have to wait for each sprint cycle to see the results. LCAP tools enable business users to see their requirements translate into software in near real-time.?The continuous engineering model of define-design-deliver-validate-optimize in near real-time brings business-IT alignment and accelerated delivery of business outcomes unmatched by other traditional development approaches.
?2.????Rapid Learning curve for developers: LCAP tools can be used by IT developers or tech-savvy business users/analysts. The learning curve to gain basic proficiency in LCAP development is in weeks as against many months for traditional technology development. As we are witnessing a major shortage of technology talent in the market, LCAP’s offer a compelling alternative to scale our delivery capacity and engage our business teams with hands-on development partnerships.
LCAP challenges, risks, and mistakes :
?I have highlighted many advantages and value points for LCAP. But LCAP tools come with their own challenges which can be detrimental to the broader adoption and success at the enterprise scale. Adopting LCAP tools without proper due diligence and governance can often lead to costly mistakes. I am sharing below some of the challenges and mistakes from my experience :
1.????Skills mismatch: Teams deployed on LCAP development need to be proficient in the respective low-code platforms with appropriate product certifications and knowledge of what to do and what not to do. Highly proficient software engineers good at writing bespoke, coding-intensive software applications may not be willing to learn and master LCAP. Or such developers may end up writing complex code with highly customized features that are difficult to maintain or scale. ?Ensuring the right mindset, skillsets, and attitude with proper investment in LCAP training and certifications are foundational to the success of solutions delivery using LCAP.
2.????Product limitations: A key mistake with low code platforms is not accounting for some of the technical limitations. In some projects, we tried to extend the low code platforms to handle high-volume, transaction-centric applications with recovery and failover features or batch processing in large volumes. We found that low code platforms do not scale and perform well in situations involving integration-centric, data or service orchestration across multiple systems or complex data structures. In such scenarios we found the inability of the LC platform to deliver the speed and quality. The applications either failed in the middle of the process or completed the process but skipped some of the records. We had no ability to ensure 100% processing of large data volumes in a batch?mode with LC platforms.?This was a major operational and regulatory challenge with a high impact on customer experience.
3.????Data, logic, and process model portability: Many low code platforms do not offer portability of data generated over time when we need to perform some analytics or visualization through other applications. Also, the lack of industry-standard modeling notations like BPMN for process modeling can make the process models non-portable when we need to migrate from one platform to another. LCAP tools offer “out of box” capabilities to develop complete applications without going through database designs (conceptual, logical, physical) or granular definitions of containers or objects. Hence, there is a certain level of sacrifice in terms of open architecture concepts while we get the advantages of speed and ease of development with a predefined configurable object model. The data is mostly stored in native databases that are inaccessible to data engineers/admins.?Hence, our ability to extract the data in desired format or filters is severely limited. LCAP tools make it difficult When we want to perform an “extract-transform-load” function to get data into third-party reporting, visualization, or data warehouses.
?4.????User interface and authentication challenges: While low code platforms make it easy to develop user interfaces, they may be restrictive in terms of screen layout, font, and navigation flexibility. Also, composite applications that require data aggregation from different sources and exposure to customers or employees outside the network, may pose challenges due to VPN configurations, virtualization software or identity and access management.
5.????Customizations and associated challenges with upgrades/migrations: LCAP tools are ideal when we don’t have to over-customize them by writing complex code to augment certain capabilities. When we over-customize LCAP tools with custom coding, we may be successful in the initial deployments but when the product vendor releases major upgrades, we may find that there is no backward compatibility.?Fixing such issues may be cost-prohibitive and throw business continuity risks.
Summary - Patterns and Anti-patterns for repeatable success with LCAP :
?LCAP tools are here to stay and offer immense value in terms of speed, cost, and agility. CIO’s and CTO’s must include LCAP as a part of the enterprise technology standards and build a governance model that will address the best practices and recommended patterns for adoption.?I am sharing the key patterns for successful LCAP adoption and anti-patterns for LCAP as below :
[Disclaimer - The views and opinions expressed within this article belong solely to me, and do not reflect the views of my current or past employers]
Enterprise Tech Arch Manager at Accenture
3 年At least for next two years LCAP is most acceptable platform.
Technology Leader | Pega | Low Code | Digital Transformation | Process Automation
3 年Well written article Vinay. A must read for anyone starting a low code application development journey.
Executive Director @ JP Morgan Chase | INSEAD |
3 年Very well written Vinaykumar Mummigatti
Top AI Voice | Founder, CEO | Author | Board Member | Gartner Peer Ambassador | Speaker | Bridge Builder
3 年Nicely done Vinaykumar.
Director, Data and AI at Wipro | Management Consulting | Enterprise Transformation | IIM Lucknow
3 年Insightful read! It's very important to evaluate and identify use cases that are fit for LCAP considering their limitations around scalability and customization. When used to solve the right business problem, LCAP is a great way to automate manual processes with minimal development effort. Thanks for sharing.