AI and automation are changing not just how work gets done, but who gets to build the systems behind it. David Walker, Chief Product Officer at Decoded, explains why agentic AI and automation are reshaping who gets to build software.
Every Operations leader I speak to has a shadow backlog.
It’s that list of small but critical tools you need to run your business efficiently. The dashboard that connects Sales to Finance. The automation that triages customer support tickets. The workflow that cleans up messy CRM data.
You’ve asked Central IT for them. You’ve been told they’re on the roadmap. Perhaps for Q3 2026.
I don’t blame IT. They’re dealing with security, infrastructure and compliance pressures that leave little capacity to build bespoke automation for every department.
For years, this created a deadlock. If you couldn’t write Java or C#, you were a passive consumer of software. You were blocked.
That barrier is starting to collapse.
We are entering the agentic shift. A fundamental change in who gets to build software.
Code is ultimately just a language. And right now, the most powerful programming language in the world isn’t Python. It’s English.
With the rise of autonomous AI agents, we’re moving from a world where syntax matters, where the semi-colon goes, to a world where logic matters. Defining outcomes, decisions and intent. Traditional coding skills are no longer the only route to building effective systems.
Stop thinking about AI as a chatbot. Chatbots talk. Agents act.
– A chatbot summarises an email.
– An agent reads the email, checks the sender against the CRM, drafts a response based on customer tier, updates the deal stage in Salesforce and pings Slack for approval.
This isn’t science fiction. Variations of this capability are already available through no-code orchestration platforms and natural language prompting. But there’s a catch. You can’t just hand these tools to an untrained Operations Manager and hope for the best.
A new architecture of work is emerging, split between pro-code developers and a growing layer of no-code architects.
1. The Pro-Code Layer: Engineering and IT
Experienced developers shouldn’t be focused on simple internal forms. Their value lies in building the engine.
– Fine-tuning models.
– Managing vector databases.
– Setting up secure API gateways.
– Defining the guardrails that keep data and systems safe.
This specialist capability remains essential.
2. The No-Code Layer: Business Context
This is where the agentic architect sits.
These are process owners, analysts and operations managers. The people closest to how work actually gets done. We already help organisations build early applied AI capability through programmes such as GenAI Champions, focused on real-world use rather than theory.
These practitioners:
– Understand the customer journey
– Know where data breaks down
– Use AI to rethink workflows, connecting systems and defining decision logic, without writing a single line of code
IT builds the roads. The business builds the cars.
This shift addresses one of the biggest challenges in software delivery. The context gap.
Traditionally, business users document requirements, pass them to developers, and hope nuance survives the handover. Something is almost always lost.
When the business expert becomes the builder, supported by AI, that gap narrows. Solutions reflect lived experience, not abstract interpretation.
The risk is assuming that natural language removes the need for skill.
Poor logic remains poor logic, even when written in English. Ungoverned agents can hallucinate, expose sensitive data or create unexpected cost and risk.
To work effectively with agentic systems, teams need new, practical skills:
– Computational thinking, to break problems into clear logical steps
– Data logic, to structure information AI systems can reason over
– Agent and workflow design, to decide where humans, automation and AI work together
– Governance, to test, evaluate and judge outputs before they reach customers
We are in the process of desiging our AI and Automation Practitioner Level 4 apprenticeship programme to build this exact hybrid professional. We aren’t training them to become coders. We are training them to become Architects of Automation.
The businesses that win in 2025 won’t be the ones with the biggest IT budgets. They will be the ones where every team has a “Local Developer”, an Agentic Architect, solving problems at the edge.
Stop waiting for Engineering. Start building Agents.
If you’re interested in finding out more as the programme becomes available, we’d love to speak with you. Get in touch to start the conversation and stay informed as this apprenticeship takes shape.
If you’d like to learn more about Decoded and how we can help transform your
organisation, we’d love to hear from you.
