Navigating Automation: How 70% Automation Is Redefining the Lifecycle of Work Without Increasing Risk

Automation isn’t just transforming industries—it’s redefining how work gets done across the lifecycle of delivery, from requirements gathering to coding and deployment. A recent case with a global IT leader revealed the potential and limits of AI-driven automation. When testing Userdoc’s AI capabilities, they found the tool could generate 70% of their user stories out of the box, significantly reducing manual effort.

But this discovery also sparks a deeper question: What should—and shouldn’t—be automated?

Automation in Action: Chris’s Perspective

Chris Rickard, founder of Userdoc, described the results of the trial:

"Userdoc's AI cut down the time to create our user stories and sped up our software development. After a thorough examination, we found the AI accuracy and relevance at around 70%—saving a lot of work."

For the IT leader involved, the remaining 30% of effort required human refinement, creativity, and collaboration. This underscores a critical point: automation is a powerful tool, but it doesn’t replace the need for human judgement.

Chris added his personal experience with AI in development workflows:

"I use GitHub Copilot 30+ times a day for coding, and would say I have similar experiences. Out of the box, it gives me 70% of what I need, and I focus on adding the 30% to refine and innovate."

What Should Be Automated?

The tasks best suited for automation are repetitive, predictable, and time-consuming. Examples include:

• Writing Requirements and Acceptance Criteria: AI tools like Userdoc streamline this process, reducing bottlenecks in delivery pipelines.

• Coding Repeated Features: Tools like GitHub Copilot eliminate the need for developers to rewrite boilerplate code.

• Work Management Updates: Automating task creation, status changes, and tracking across tools reduces admin overhead.

By automating these, teams can focus on high-value, creative work.

What Shouldn’t Be Automated?

Despite its potential, automation has its limits. Tasks that require creativity, collaboration, and nuanced understanding should remain human-led. These include:

• Product Discovery: Customer interviews, user research, and visioning workshops rely on empathy and insight that machines can’t replicate.

• Innovation Workshops: Collaborative sessions thrive on human interaction and spontaneous idea generation.

• Multi-Stakeholder Alignment: Negotiating competing priorities and defining a shared vision require human connection.

While automation can augment these processes, it shouldn’t replace them.

Bridging the Gap Between Automation and Human Expertise

The key to successful automation is balance:

1. Automate Repetitive Tasks: Save time and free up teams for more meaningful work.

2. Augment High-Value Work: Use AI as a tool to enhance creativity and speed without compromising quality.

3. Build Seamless Workflows: Ensure that AI tools integrate smoothly into existing processes, empowering—not overwhelming—teams.

As Chris Rickard noted, the goal of automation isn’t to replace human effort but to focus it where it matters most:

"Automation is about enabling people to spend their time on high-value work, not just reducing costs or speeding up timelines."

Final Thought: Automation Without Risk

Automation has the power to redefine work across the lifecycle—from requirements definition to delivery. But its success depends on what you automate, how you implement it, and how you balance it with human expertise.

By automating 70% of predictable tasks and reserving 30% for human refinement, teams can unlock both speed and quality—without increasing risk to their processes or tech stack.

Curious how automation could optimise delivery for your teams? Let’s compare notes and explore the possibilities together.
Book a time with me