The Future of Software Development is Visual

At Contextual, we have a fundamental belief about the future of software development: with the assistance of AI, development will increasingly be done—not by traditional software developers—but by subject matter experts, business analysts, or any business stakeholder who has an idea and wants to bring it to life.

This future of AI-assisted development will be visually driven, allowing non-technical users to design and orchestrate solutions without needing to write code. Visual orchestration will be the key to making software development accessible to everyone, from experts to complete beginners.

The road to realizing this vision is built on a few key concepts that are helping us to accelerate the delivery of what we call "SolutionAI" — a framework that enables both developers and non-developers to build software solutions quickly and easily, without deep programming knowledge.

Building Blocks are Critical

Today’s AI assistants (e.g., GitHub Copilot) for developers are fantastic at certain things, like generating small code snippets or completing functions. However, the limitations of these AIs often relate to the scope and complexity of what they're asked to do. While AI tools like OpenAI's models are improving, they're most effective when tasked with discrete functions or methods rather than entire systems or services.

The transition from today’s reality—where AI assists in creating small parts of software—to a future where AI can build entire systems, requires a framework that simplifies the problem for AI.

Our Approach: Reducing Complexity

When AI is asked to solve highly complex, ambiguous tasks (e.g., “build an IoT framework for receiving and processing events from thousands of devices”), it often struggles due to the number of unknowns involved. The AI needs clarification on network setups, technology choices, device types, data processing methods, and end-user requirements before it can even start.

This is why context and constraints are so crucial. The more you can reduce the complexity and scope of the task, the better AI can perform. By breaking down large problems into smaller, well-defined tasks with clear boundaries, we give the AI a fighting chance to produce meaningful, usable outputs.

At Contextual, we reduce this complexity by providing a visual orchestration platform based on discrete building blocks. Our platform, which extends on the visual power of Node-RED, leverages a data layer based on JSON-schema that automatically provides APIs, UX components, and triggers to send data to flows. By limiting the set of tools available to solve a problem, we reduce the problem space and empower AI to generate solutions effectively.

Data Models Must Be Well-Defined

One of the key enablers of AI in software development is the need for well-defined data models. Without a clear understanding of how data is structured and flows through the system, even the most advanced AI can struggle to make sense of a complex problem. At Contextual, we ensure that our data models are built around standardized JSON-schema, making them easy for both humans and AI to work with.

Service-Oriented Architectures Can Be an Enabler

Another key aspect of simplifying software development is the move toward service-oriented architectures (SOA). Software has already begun moving from monolithic systems to modular services that can be easily composed, shared, and reused across different applications. This shift will enable cross-system functions and more accessible shared data sets, reducing the need for traditional development and promoting flexibility and scalability. Further, it aligns well with AI-assisted development because the component services are modular, with defined boundaries that the AI can easily understand.

A Full Lifecycle Approach is Required

AI must be involved throughout the software lifecycle, from development to debugging and maintenance. Our SolutionAI for Operations enhances problem detection and resolution with AI-assisted debugging. By analyzing logs, AI identifies errors, suggests fixes, and traces root causes, especially in distributed systems.

AI-Driven Log Analysis

AI-powered log analysis is key to understanding system behavior. It correlates logs across services, proactively detects anomalies, and summarizes vast amounts of data for quick insights, saving developers from sifting through raw logs. This allows teams to resolve issues more efficiently and avoid potential failures before they escalate.

Continuous Improvement

AI doesn’t just fix issues—it learns from them, enabling continuous improvement. Over time, AI can suggest code optimizations or redesigns to prevent recurring problems. AI also plays a key role in post-deployment monitoring, ensuring systems are reliable and scalable.

The Development Experience Will Always Include AI

In just a few years, AI has become a fundamental part of the development process. Tools like GitHub Copilot, which was launched in 2021, and OpenAI’s ChatGPT (launched in 2022), have already transformed how software is written. Today, 92% of US-based developers are using AI in some capacity to assist them in their jobs¹, and 46% or more of all new code is generated by AI assistants when developers use such tools². The future of development is, and will continue to be, heavily influenced by AI integration.

Human Understanding is a Requirement

As AI becomes more involved in creating software, it is critical that humans can still understand the systems being built. For example, when a business analyst collaborates with an AI assistant to create a software solution, that analyst is still responsible for ensuring the solution meets the business’s needs.

This means the analyst must understand how the solution works, verify the AI’s tests, and collaborate with the AI to refine and improve the software. In this context, human understanding is not just a requirement—it’s an essential part of the development process. This is why visual orchestration is so critical to mass adoption.

The Future is Visual

The easiest way for humans to understand AI-developed solutions is through visual representations. Visual programming, a concept that has been around for decades, offers an accessible way to represent logic and workflows. Early tools like Scratch and Blockly aimed to lower the barrier to entry by allowing users to build programs visually using blocks and drag-and-drop interfaces.

However, while these early tools were great for teaching programming, they lacked the power to create complex, real-world software solutions. At Contextual, we are building on the foundation provided by such tools to create a visual programming environment that can be used not only for learning but for real-world applications.

Characteristics of a Good Visual Programming Framework

A good visual programming environment, whether used by a human or an AI, should:

• Be intuitive and easy to use
• Clearly visualize the flow of logic
• Limit the building blocks available to users to avoid overwhelming complexity
• Provide standard solutions for common tasks
• Allow custom coding when needed
• Support event-driven programming and parallel execution

These same characteristics make a visual environment ideal for collaboration between human users and AI developers. By presenting software in a clear, visual way, both human stakeholders and AI assistants can work together to build robust solutions.

At Contextual, we believe that the future of software development is visual, modular, and AI-driven. Our platform provides the tools necessary to make this future a reality, empowering users of all skill levels to build, orchestrate, and manage AI-powered software solutions without needing to write code.

‍