Building an AI Operating System for India’s Construction Sector

For years and still, India’s construction sites have run on registers, WhatsApp groups, and fragmented spreadsheets. Daily records of material usage, labour attendance, machinery deployment, and project progress often remain scattered across notebooks and photos shared between site managers.

That’s what a Bengaluru-based construction tech startup wants to fix. Powerplay builds AI agents specifically designed for contractors managing large infrastructure and real estate projects.

Founded in 2021 by IIT Roorkee alums Iesh Dixit and Shubham Goyal, Powerplay started by digitising site-level workflows for mid-sized contractors. Today, the company is evolving into what Dixit describes as a proactive AI layer for construction operations.

“We thought, can we build something which helps them act very proactively and make sure that the right things happen at the right time so that the work keeps on going,” Dixit tells AIM.

The startup has raised around $15 million to date from investors including Accel, Peak XV Partners, and India Quotient.

Replacing Registers With Software

Before starting up, Dixit had a short stint as a software developer. His first go at entrepreneurship was in the B2B commerce space before joining retailtech startup Jumbotail. He credits those years with helping him transition from a product-focused mindset to a business-oriented approach.

He then later began researching operational inefficiencies in construction workflows, which led him to found Powerplay.

Powerplay’s initial target users were neither large EPC (engineering, procurement, and construction) giants nor very small contractors. Instead, the startup focused on the middle layer of contractors handling projects worth ₹10–20 crore and above and companies managing multiple sites with lean back-office teams.

According to Dixit, the original problem was not the absence of ERP systems in offices, but collecting reliable data from construction sites.

“Everyday material is coming, labour is coming, machinery is being used. All of this data is on the site, on a register, or sometimes people click photos and put them on WhatsApp groups,” the founder and CEO explains.

Powerplay’s first product digitised this site-level information through a lightweight app designed specifically for field workers. Instead of charging heavily upfront, the company distributed the app for free to drive adoption.

The strategy helped the company scale rapidly. Currently, around 40,000 businesses use Powerplay’s app ecosystem every month.

Building AI Agents for Procurement and Finance

Over the last year, the company began layering AI capabilities on top of the operational data flowing through its platform.

The startup observed that many contractors manage crores worth of procurement and operations with back-office teams of just 10–15 people. That became the trigger for building role-specific AI agents.

Powerplay now offers procurement agents that analyse work requirements, inventory levels, and existing orders to recommend what materials should be purchased. Finance-focused agents help teams prioritise bill clearances, while project planning agents can interpret drawings and estimate resource requirements.

The system combines NLP (natural language processing) models, orchestration layers, and retrieval-based querying systems to route user requests and generate responses using operational project data.

“We bundled it with the software. There are credits now. People can use those credits and use the agents,” Dixit says.

An AI-Heavy Team

Despite serving thousands of businesses, Powerplay operates with a relatively lean engineering setup. The company has around 100 full-time employees overall, but only 13 people in engineering. Dixit says AI coding tools now play a major role in development workflows.

“My spending on AI coding tools is equivalent to a person's salary in the engineering team,” he compares.

The startup also employs three AI-native engineers focused specifically on building agentic workflows and AI products. Meanwhile, developers increasingly spend more time reviewing and quality-checking outputs rather than manually writing code.

“We are aspiring towards a setup where developers manage agents and get work done,” he adds.

The India Pricing Challenge

While AI tooling startups globally often charge based on token usage or software consumption, Dixit believes Indian enterprises operate differently. He believes Indian customers are significantly more outcome-driven and cost-sensitive than their global counterparts, pushing startups to rethink traditional AI pricing strategies.“A lot of people don’t believe in making money from technical products in India,” he reflects.

Instead of relying purely on SaaS-style pricing, Powerplay combines subscription models with outcome-linked commissions. For example, when procurement agents successfully help contractors source materials through the platform, the company earns a percentage of the transaction.

However, the startup has refused to disclose its topline.

Powerplay operates in a growing construction-tech ecosystem—growing at 14.3% and forecast to reach $1.26 billion by 2030, according to GrandView Research—alongside startups such as Site Setu, Yojo, and FalconBrick.