At 30% time and effort savings, AI coding is worthwhile

With automation and AI, coding time and resources can be shrunk in a compelling way. But what about test-flakiness, patterns, regression testing, test coverage, robustness, redundancies and developer satisfaction? Richard Spence, Area Vice President of Growth Sales, International at UiPath decodes all that and more.

- Pratima H

What new advantages apart from cost-cuts, time-cuts and automation does agentic AI bring in the realm of software testing?

Agentic AI is reshaping the future of testing. Some of the direct advantages are in quality, savings and productivity. Infusing quality at the requirement stage ensures clear, testable requirements. Innovative test ideas and mapping requirements to test cases enhance validation and traceability, driving continuous improvement for robust software. When it comes to effort saving, we can generate test cases from requirements in any format (descriptions, pictures, wireframes). We can now drive automation test cases and review bulk results with actionable insights, saving effort and enhancing efficiency.

And that ties into productivity?

Productivity booster agents achieve huge efficiency gains with ready-built agents and the ability to create custom agents using our agent builder, streamlining processes and enhancing productivity.

Developers tend to struggle with real efficiency gains, waiting-time-on-builds-and-tests, redundancy-fears and new skills when it comes to low code and Al-code. Your take?

Today teams are overwhelmed by numerous options and social feeds. While many vendors tout AI-based testing, true value realisation is something that needs to be looked at closely. In my view, basic capabilities of a real agentic testing should help us drive huge efficiency gain ensuring quality to avoid retesting.

How?

Adapting to changing priorities, create a lot of reutilisation factors to address resource availability. Simple capabilities like Shift-left testing to ensure early test case creation, wireframe utilisation, strong priority mapping, and flexibility to integrate with existing tool chains and skillsets, low/no/pro code options and purpose-built agents should deliver tremendous value for the team.