A senior Infosys executive called for urgent investment in worker training and data plumbing as companies rush to apply artificial intelligence. Rajan Padmanabhan, an associate vice president and Unit Technology Officer for Data Analytics and AI at the Indian IT giant, stressed that skills and clean data are now make-or-break factors for enterprise AI programs.
His message comes as businesses across sectors scale pilots into production and face real-world hurdles. Firms want returns this year, not in a distant future. Many are learning that staff readiness and trustworthy data pipelines decide whether deployments succeed or stall.
Why Skills Lag Behind AI Ambitions
Companies have spent the past two years testing large models and automation. Now they need people who can manage prompts, govern models, and measure value. That includes engineers, product owners, and frontline teams who use new tools daily.
The World Economic Forum has warned that millions of roles will change as automation spreads, pushing reskilling to the top of executive agendas. Consulting firms also report a sharp rise in demand for training on model safety, MLOps, and data stewardship.
Rajan Padmanabhan, AVP and Unit Technology Officer for Data Analytics and AI at Infosys, highlights the importance of upskilling and data infrastructure
His framing reflects a common lesson from early projects. Tools move fast, but workforce habits change slowly. Without shared standards and repeatable playbooks, teams struggle to scale good ideas across business units.
Data Infrastructure as the Bottleneck
Even the best teams fail when data is messy, stale, or locked away. Many firms still rely on legacy systems and manual extracts. That creates risk and drags down model performance.
Analysts have found that most AI effort sits upstream of modeling. The hard work is mapping data, defining ownership, and building pipelines that capture consent and lineage. When these parts are weak, accuracy drops and trust erodes.
Padmanabhan’s focus on data plumbing aligns with this reality. Clean inputs and clear controls help scale use cases across sales, finance, and operations. They also reduce duplicated work and audit headaches.
What Companies Can Do Now
Enterprises that move from pilots to value share a few habits. They target problems with near-term payback. They train people closest to the work. And they make data quality visible to leaders.
- Launch role-based training for engineers, analysts, and business users.
- Stand up a small, empowered AI governance group with clear decision rights.
- Invest in data catalogs, lineage tracking, and access controls.
- Tie projects to metrics such as cycle time, win rate, or cost per case.
- Create a reuse library for prompts, templates, and verified datasets.
These steps build momentum without large bets. They also help avoid shadow projects that create risk or waste money.
Balancing Optimism with Caution
Many executives see rapid gains in coding help, support chat, and document search. Yet there are concerns. Legal teams fear leaks of sensitive data. Compliance groups worry about bias and traceability. Finance leaders want clear returns before expanding spend.
Industry voices differ on pacing. Some push fast adoption to stay competitive. Others prefer slower rollouts tied to sharper guardrails. Both camps agree on one point: skilled people and solid data are non-negotiable.
Signals to Watch
Hiring patterns are a leading indicator. Growth in data engineering and governance roles signals a shift from hype to scale. Budget moves also matter. Spend on data platforms often precedes broader AI rollouts by one to two quarters.
Regulation is another driver. New rules on transparency and model risk will raise the bar for controls and documentation. Firms with mature data pipelines will adjust faster and at lower cost.
Padmanabhan’s emphasis on “upskilling and data infrastructure” captures the crux of this moment. The tools are here. The winners will be those who teach their people well and fix their data first. Watch for companies that invest early in training and build strong data backbones; they are most likely to turn pilots into profits this year.
