Software engineers in the AI age: Skills beyond coding

For decades, learning to code has been considered a pathway to a career in the technology industry. Today, artificial intelligence (AI) tools can generate code, identify errors and explain programming concepts in seconds. This has prompted a question among students, parents and educators: as AI takes on more coding tasks, what skills will future software engineers need?

Educators argue that software engineering is evolving rather than disappearing. While AI can automate parts of development, building reliable software still requires problem-solving, technical understanding, system design and the ability to evaluate the quality of a solution.

Beyond writing code

AI-assisted coding is changing how developers approach routine tasks. Instead of writing every line manually, engineers can use AI tools to generate initial code, suggest fixes and explore different approaches. This can reduce time spent on repetitive work and allow greater attention to complex problems.

However, generated code must be reviewed before it is used. Developers need to assess whether it is secure, efficient, maintainable and suitable for the intended application. They must also identify errors that may not be immediately visible.

The shift is comparable to the introduction of calculators in mathematics. Calculators reduced the need for repetitive arithmetic but did not eliminate the need to understand mathematical concepts. Similarly, AI can assist with coding without replacing the engineering knowledge required to build dependable systems.

Viraj Sagar Das, Pro Chancellor at BBD University, emphasises that software engineering involves more than typing instructions into a computer. Developers must understand what needs to be built, how different components interact and how systems should respond when something goes wrong.

Why fundamentals still matter

Prompting allows users to communicate instructions to AI systems, but effective prompts alone cannot replace a foundation in computer science. Algorithms, data structures, databases, cybersecurity, computer networks and system design help developers understand the problems they are trying to solve.

Educators stress that students need to learn how to use AI as an assistant while retaining the ability to reason independently. Without technical fundamentals, a developer may struggle to recognise incorrect suggestions or understand the consequences of a design decision.

AI literacy is therefore becoming an additional skill rather than a substitute for programming knowledge. Future professionals will need to know which tasks can be delegated to AI, how to verify its responses and when human judgement is essential.

Rethinking software education

The growing use of AI is also raising questions about how programming students should be assessed. If a tool can produce a complete assignment, the final submission alone may not reveal how well a student understands the underlying concepts.

Educators are considering assessment methods that focus on evaluating AI-generated code, finding vulnerabilities, debugging unfamiliar programs and explaining technical decisions. Designing complete systems and defending the reasoning behind a solution can provide a clearer picture of a student’s understanding.

At BBD University, AI components have been incorporated into curricula, alongside exercises in which students explain code and debug programmes they did not write themselves. Such approaches aim to assess understanding and problem-solving rather than code production alone.

What lies ahead for India’s technology workforce?

For India, which has a large software and IT services sector, these changes have implications for engineering education and recruitment. Employers may increasingly look for graduates who can combine programming knowledge with AI-assisted development, critical thinking and responsibility for the systems they build.

The pace and impact of change will vary across roles and organisations. AI may reduce the time needed for some tasks while creating new requirements in areas such as verification, integration, security and system oversight.

The software engineer of the future may write fewer lines of code manually but will still need to understand how that code works and whether it solves the right problem.

The central challenge for universities is not simply to teach students how to prompt AI tools. It is to prepare them to use those tools thoughtfully, question their outputs and take responsibility for the final result. In an AI-assisted industry, engineering knowledge remains essential to turning generated code into reliable software.

(The author is Executive Director at COER University)

ABOUT THE AUTHOR

Dr Charu Jain