Iesh Dixit, CEO & Co-Founder at Powerplay in conversation with Efficient Manufacturing Magazine
Q. How do you see AI-led workflow automation transforming the traditionally manual and fragmented construction industry over the next 3–5 years?
Over the next three to five years, AI-led workflow automation will push construction away from reactive project management and toward something much more proactive in day-to-day execution. Right now, most sites still depend heavily on manual coordination – calls, WhatsApp threads, spreadsheets, and scattered records. Information exists, but it’s rarely in one place, and decisions slow down because people are working with partial visibility.
AI starts to change that by turning operational data into workflows that actually guide action in real time. Instead of recording what already happened, systems begin to surface risks early – delays, cost drift, and missed dependencies and suggest the next move before the issue gets bigger.
In practical terms, the shift is from people holding everything in their heads to systems carrying the operational load. Site supervisors will lean on AI agents to handle routine planning, trigger procurement at the right moment, and keep progress tracking consistent. Financial reconciliation, which today often drags on for weeks after a project phase ends, becomes something that happens continuously in the background.
Q. Construction has long struggled with delays, cost overruns, and talent shortages. How can vertical AI solutions help address these systemic challenges at scale?
At the core, construction doesn’t suffer from a lack of effort or skill. Most teams work incredibly hard. The real problem is coordination. Materials, labour, timelines, and finances all move at different speeds, and keeping them aligned is inherently messy. Vertical AI is useful here because it’s built around the actual workflows of the industry, not generic productivity tasks. Take delays, for example. They often happen because dependencies aren’t visible in real time. A vertical system can track progress continuously and trigger the next requirement – materials, labour, equipment – before the bottleneck shows up.
Cost overruns usually trace back to poor visibility or unplanned spending. If budgets and actual expenses are reconciled continuously, deviations surface early. Not weeks later, when the damage is already done. Talent shortages are another pressure point. A lot of experienced supervisors spend their time on repetitive coordination instead of higher-value decisions. When AI agents handle routine tracking and planning, those professionals can focus on judgment, quality, and problem-solving. At scale, the effect compounds. Instead of hiring more people to manage complexity, companies standardise execution through systems. That’s when problems that once felt structural start to look operational and manageable.
Q. With AI increasingly moving from assistive tools to execution-led platforms, do you believe construction could become one of the biggest adopters of domain-trained AI in India? Why?
Yes, and in many ways, construction is unusually well-positioned to adopt domain-trained AI at scale in India. The reason is fairly straightforward. The industry has a large productivity gap, and it runs on a constant stream of operational decisions. Every project generates hundreds of small choices every day – what to buy, where to deploy labour, how to sequence work, how to manage cash. Those decisions are repetitive, but they carry financial consequences.
That makes construction a natural fit for execution-led AI. Unlike sectors that are already heavily digitised, construction still carries a lot of manual overhead. That creates a clear payoff for automation. Even modest improvements in planning accuracy or cost control can shift project economics in a noticeable way. There’s also a macro factor. India’s construction market continues to expand due to urbanisation, infrastructure programmes, and housing demand. As project volumes rise, informal coordination starts to break down. Standardised execution systems stop being optional. In that environment, domain-trained AI won’t feel like an advantage reserved for a few companies. It will look more like basic infrastructure.
Q. Powerplay’s platform is built on insights from over 85,000 construction projects. How has this deep industry data helped shape the development and accuracy of your AI agents?
The scale and diversity of project data matter more than people sometimes realise. Construction workflows vary widely – project type, geography, contractor size, and execution stage. Without enough historical exposure, automation tends to stay generic.
By analysing patterns across more than 85,000 projects, we’ve been able to see recurring behaviours – how materials get consumed, how schedules shift under pressure, how costs evolve, and where delays tend to originate. Those patterns repeat more often than they appear to on the surface. That history allows AI agents to operate with context rather than rules alone.
For instance, if one activity slips, the system can anticipate the downstream effect on procurement or labour deployment because it has seen similar sequences before. If spending starts drifting from expected norms, the system can flag it early, not because of a static threshold, but because the pattern looks familiar. In a practical sense, the data gives the system memory. And that memory is what builds confidence. Accuracy, relevance, and trust are non-negotiable in construction. Without those, adoption stalls quickly.
Q. As AI becomes more deeply integrated into construction workflows, how do you envision Powerplay shaping the future of project execution from planning and procurement to on-site decision-making and financial reconciliation?
Our view is that project execution should feel continuous, connected, and largely self-managing. Today, workflows are fragmented. Planning tools sit in one place, procurement processes in another, site operations somewhere else, and financial records in yet another system. Each function works, but rarely in sync. That gap creates delays and forces teams to reconcile information after the fact.
We see Powerplay evolving into a single execution layer that ties those functions together.
At the planning stage, AI can generate structured schedules and resource plans using project scope and historical benchmarks. During execution, agents monitor progress in real time, trigger procurement automatically when thresholds are reached, and adjust timelines as conditions shift on the ground.
Financially, reconciliation becomes ongoing rather than periodic. Expenses, invoices, and budgets stay aligned as the project moves, so stakeholders always have a clear picture of financial health. The long-term aim is stability. When systems can anticipate issues, coordinate resources, and maintain financial discipline on their own, teams gain room to grow without losing control.
