Business growth worldwide is increasingly influenced by data-driven innovation. In Kenya, both private and public institutions are gradually adopting generative AI to gain more control over their processes and data. This trend is especially notable in application development. Generative AI (GenAI), powered by large language models (LLMs) trained on billions of coding parameters, assists developers by generating quick solutions for repetitive tasks such as boilerplate code, database operations, and standard UI elements. This leads to a faster and improved application development experience.
Can GenAI Replace Low-Code Development Platforms?
The simple answer is no. Even before GenAI’s entry into app development, low-code and no-code development platforms were already simplifying development environments. They helped businesses achieve agility in software development life cycles and go-to-market strategies.
While GenAI cannot entirely replace low-code platforms without human involvement, it can significantly enhance their value. For instance, GenAI can generate specific blocks of code based on a prompt. However, AI lacks the ability to determine where to insert these snippets, how tweaking components affects outcomes, or if the snippets can be optimized for better results.
The best approach is to select a low-code/no-code platform with strong GenAI integration and robust LLM capabilities. A loosely coupled low-code/AI approach could lead to technical debt, poorly designed applications, and compliance challenges.
Key Considerations for Using Low-Code Platforms with AI
- Platform Maturity: Avoid choosing a low-code platform solely for its AI capabilities, as this technology is still evolving. Instead, focus on mature platforms with comprehensive features to support the creation of scalable custom applications.
- Privacy and Security: Applications built on low-code platforms often interact with various types of data. It is crucial to address how data will be used within the context of LLM upfront. Choose platforms that prioritize data privacy and security in all their provisions, including LLM capabilities.
- Compliance: Non-compliance can lead to severe financial penalties and loss of customer trust. Select platform vendors that adhere to major regulations in their operating regions. For example, a platform should monitor application progress and flag GDPR non-compliance if the app is meant for the EU market.
- Governance: Effective governance measures are essential, especially when different developer profiles use the platform to solve business problems. LLM-based capabilities can compound governance issues, as users might integrate foreign code blocks into the existing code base.
Approach to Adoption
In the future, reputable long-term low-code platforms will train and launch proprietary LLMs to better control outputs. Contextual, domain-specific LLMs will also be deployed with low-code platforms to build industry-specific or use-case-specific applications at scale.
It is important not to jump into new technology due to hype. Instead, adopt a careful approach: start small, assess progress, and then scale. Comprehensive onboarding interventions can help balance abstraction and control within low-code platforms.
By considering these factors, businesses can leverage the strengths of both low-code platforms and GenAI to enhance their application development processes.
