E-Paper
From Writing Code to Supervising AI: Engineering enters a new era
Agentic systems may automate repetitive coding and operational tasks, but humans must still define goals, ask better questions, interpret results among others.
Coding fluency still matters, but it no longer defines engineering readiness. Agentic AI now demands engineers who can design intelligent systems that reason, plan, use tools, coordinate workflows, and operate safely under human supervision. Unlike chatbots that mainly respond to prompts, AI agents can take a goal, break it into steps, retrieve information, test outputs, monitor progress, and improve their approach. They may inspect code, suggest fixes, run tests, summarize documents, route requests, and escalate exceptions. The engineer’s role is shifting from writing every instruction to supervising the larger system.
The Workforce Signal
The workforce signals are clear. The World Economic Forum’s Future of Jobs Report 2025 projects 170 million new jobs and 92 million displaced jobs by 2030, a net gain of 78 million roles. It also identifies AI and information-processing technologies as major forces reshaping business models. The Agentic AI market is projected to grow from USD 7.84 billion in 2025 to USD 52.62 billion by 2030, at a CAGR of 46.3%. AI is not just replacing work; it is changing the work people are valued for.
What every parent should know about CBSE's three-language policy
Where Agentic AI Is Already Emerging
Agentic systems are moving from pilots to practice. In banking, they support service, fraud triage, loan processing, and compliance. In healthcare, they assist with intake, documentation, claims, and care coordination. In legal, corporate, and education settings, they review contracts, summarize obligations, personalize learning, generate practice tasks, and reduce routine workload. The shift is now entering everyday professional work.
Why Coding Alone Is Not Enough
Coding alone cannot define future readiness. An AI agent is useful only when its goal, data, tools, boundaries, and evaluation criteria are well designed. A healthcare agent must protect privacy and safety; a financial agent must follow regulations; a multilingual agent must handle accents, dialects, and code-mixed language. The challenge is not just to build a working system, but one that can be trusted.
Turning international education into career capital for Indian graduates
Skills for the Agentic AI Era
Engineers still need programming, data structures, cloud platforms, machine learning. They must also understand how components interact, where failures spread, and how human decisions fit into automated workflows. Future-ready engineers should design guardrails, observability, fallback mechanisms, audit trails, and human-in-the-loop review from the start.
Ethical judgment is equally important. Agentic systems may optimize narrow goals while missing fairness, privacy, safety, or social context. Engineers must understand responsible AI, bias detection, data governance, security risks, explainability, and regulation. The future engineer is not only a builder, but also a reviewer, risk assessor, and steward of technology, especially when AI influences hiring, credit, healthcare, education, or public services.
From Policy to Practice: Can schools deliver on the third language vision?
Interdisciplinary problem framing will matter too. Real problems involve business goals, users, institutions, domain knowledge, culture, and trade-offs. The most employable engineers will be T-shaped: deep in one technical area, yet broad enough to work with domain experts, product managers, designers, legal teams, and users. They must know what to build, why it matters, who may be affected, and how success should be measured.
Implications for Education and Upskilling
Education and industry must respond together. Programming exercises remain useful, but they should not be the endpoint. Students need project-based learning where they design AI-enabled systems, evaluate outputs, test failures, and reflect on ethical implications. Internships, live projects, and upskilling can connect theory with organizational realities. As agentic AI evolves, the ability to learn, unlearn, and relearn may become as valuable as any single technical skill.
The Human Edge
Agentic systems may automate repetitive coding and operational tasks, but humans must still define goals, ask better questions, interpret results, judge risks, and take responsibility for outcomes. The human edge lies in creativity, context, empathy, ethical reasoning, and strategic judgment.
The future workforce will not be defined by how quickly people type code, but by how wisely they guide intelligent systems toward useful, safe, and inclusive outcomes. Coding will remain a foundation, but the higher-value skill will be orchestration: combining technical fluency with systems thinking, domain understanding, and responsible judgment. Engineers who master this shift will not be displaced by AI; they will direct it with purpose.
(This article is written by Dr. Chetana Gavankar, Professor in the Work Integrated Learning Programmes (WILP) and serves as the Program Lead for the CSIS department at BITS Pilani)
