Eshan Jayamanne, Founder and CEO of Krane – Interview Series

Eshan Jayamanne, Founder and CEO of Krane, combines deep construction industry expertise with a background in engineering, analytics, and technology. Prior to founding Krane in 2023, he worked on major projects for organizations including Microsoft, Chevron, UCSF, and Caltrans, focusing on productivity improvement, supply chain optimization, lean construction, and operational efficiency. His experience managing complex construction projects exposed the industry’s persistent coordination and procurement challenges, inspiring him to build technology that helps teams make better decisions and reduce costly delays.

Krane is an AI-powered construction supply chain platform designed to improve visibility and coordination across procurement, materials management, and project operations. The platform uses AI agents to automate tasks such as material tracking, supplier management, risk identification, and invoice reconciliation, helping construction teams reduce delays and improve efficiency. Focused on large-scale projects such as data centers, healthcare facilities, and infrastructure developments, Krane aims to modernize one of the construction industry’s most fragmented and data-intensive workflows.

You spent years working across construction engineering, production analytics, project optimization, and large-scale infrastructure projects before founding Krane. After seeing firsthand how procurement delays, material tracking issues, and fragmented workflows affected major projects, what was the specific moment that convinced you AI could solve problems the construction industry had struggled with for decades?

I spent more than a decade working on complex energy, infrastructure, and data center projects, including leading projects for companies like Microsoft and Chevron. No matter the project size, I kept seeing the same challenges: teams were inundated with submittals, materials were released late, and deliveries weren’t coordinated correctly, creating costly delays.

What convinced me AI could make a difference was seeing how much of this work was still being managed through disconnected spreadsheets, emails, and phone calls. Construction teams weren’t lacking expertise; they were spending too much time chasing information across fragmented workflows.

Recent advances in AI made it possible to read construction specifications, drawings, schedules, and procurement records, then automate many of the coordination tasks that traditionally required significant manual effort. That’s when it became clear that AI could help shift the operational burden off project teams and give them more time to focus on building.

That idea ultimately became Krane. Today, the platform manages more than $17 billion in active construction projects across North America and recently raised a $9 million Seed round to continue expanding our AI-native construction operations crew.

Construction has historically lagged behind industries like finance and software in technology adoption. Why do you believe the current wave of AI is different, and why is the industry finally ready for more autonomous systems?

The industry is facing two challenges at the same time: increasing project complexity and a growing labor shortage. The demand and complexity of data centers, healthcare facilities, and other large-scale projects are putting more pressure on construction supply chains than ever before. At the same time, experienced professionals are retiring faster than they’re being replaced.

Tools like ChatGPT also helped demonstrate what AI could actually do in a practical, user-friendly way. That made the technology feel much more accessible and accelerated adoption of AI solutions built for specific industries, including construction.

Today, construction teams are being asked to deliver larger and more complex projects with less margin for error while managing volatile and difficult-to-predict supply chains. That’s why we’re seeing so much interest in AI. The industry isn’t adopting technology for technology’s sake; it’s looking for practical ways to help teams manage complexity, maintain visibility, and keep projects moving.

Krane describes itself as an AI-native platform for managing construction supply chains. What are the biggest inefficiencies that still exist today in procurement, submittals, deliveries, and supplier coordination that most people outside the industry don’t realize?

Most people are surprised by how much of construction procurement still runs through spreadsheets, emails, phone calls, and manual follow-up. A large project can involve hundreds of materials, dozens of suppliers, thousands of documents, and constantly changing schedules, yet many procurement decisions are still made based on vendor estimates, historical experience, and assumptions about market conditions.

As we looked across the projects running on Krane, we found that many of the most costly schedule delays and budget overruns could be traced back to procurement decisions made months earlier. The challenge was identifying risks early enough to do something about them.

In our own data, procurement risks surfaced an average of 47 days before they appeared on a project schedule. That means teams can have weeks of exposure to a risk before it becomes visible enough to take action.

That’s exactly why we launched Procurement OS. Krane’s new Procurement OS module brings live supply chain intelligence directly into procurement workflows. Because Krane already manages the day-to-day flow of materials across more than $17 billion in active construction projects, we’re able to capture real lead times, supplier performance, submittal cycle times, and material availability as they happen.

Procurement OS automatically identifies critical materials from project specifications, benchmarks live lead times, scores suppliers based on actual performance, and screens bids for compliance before they’re awarded. We’re moving construction from procurement by assumption to procurement by evidence.

Many AI companies focus on generating insights, while Krane appears to be moving toward AI agents that can actively execute tasks. How do you see the role of AI evolving from assistant to operator within construction workflows?

We believe the next evolution of AI in construction is moving from providing information to helping teams execute work. Construction projects generate an enormous amount of data, but identifying an issue is only the first step. Teams still need to coordinate suppliers, validate lead times, manage submittals, and keep materials moving.

That’s why we’ve built an AI construction crew where each agent owns a specific part of the materials workflow.

• Milo turns specs, drawings, and schedules into complete submittal and procurement logs in minutes
• Arlo connects siloed schedules, submittals, and procurement logs into one system and provides a clear procurement strategy for the pre-construction teams
• Chase follows up with trade partners to prioritize submittals, validate lead times, and material statuses
• Lana creates submittal templates and QCs so that GCs and trade partners can improve the quality of submittal packages and submit them on time
• Rio automates delivery scheduling, prevents any conflicts, and analyzes equipment usage
• Theo controls the procurement process from quote comparison, RFP creation to delivery

Humans still make the important decisions, but AI handles the work that historically consumed hours of a project team’s day.

Data centers have become one of Krane’s major focus areas. As AI demand drives a global construction boom for digital infrastructure, what new supply chain challenges are emerging that traditional project management tools struggle to handle?

Data centers are exposing the limits of traditional procurement processes.

Many of the critical components that power these facilities now have extraordinarily long lead times. Power transformers can take three to five years to procure, while medium-voltage switchgear can exceed 60 weeks. At the same time, electrical and mechanical equipment can represent up to 75% of a project’s guaranteed maximum price, which means procurement decisions carry enormous financial consequences.

Traditional project management tools are designed to track project status after decisions have already been made. They don’t provide real-time visibility into supplier performance, shifting lead times, or emerging supply chain constraints.

Procurement OS addresses these challenges. By leveraging live supply chain data from active projects, teams can make procurement decisions based on what’s happening in the market today rather than what happened on the last project.

One of the biggest concerns around AI agents is trust. In construction, where delays or mistakes can have multi-million-dollar consequences, how do you balance automation with human oversight?

Trust is critical because construction decisions have real schedule and budget consequences.

Our philosophy is that AI should handle coordination, while people remain responsible for judgment and decision-making. Every AI action in Krane includes a confidence score, critical decisions require human approval, and activity is logged through audit trails. That gives teams visibility into what the system is doing and why.

The other side of trust is results. Across our portfolio, Krane has tracked more than 2,000 supply chain risks before they reached the jobsite and helped customers achieve a 92% on-time delivery rate. When teams consistently see risks identified earlier and coordination happening faster, trust naturally follows.

Construction projects often involve dozens of contractors, subcontractors, suppliers, and stakeholders working across disconnected systems. How difficult is it to create a true single source of truth, and what role does AI play in making that possible?

It’s one of the hardest problems in construction.

The average project involves information spread across multiple stakeholders and systems. General contractors, subcontractors, owners, suppliers, and consultants are all working from different data sources, which creates gaps, conflicting information, and delays.

AI plays an important role because it can connect and interpret information across those environments. Krane brings together schedules, drawings, procurement logs, submittals, deliveries, and supplier communications into a single environment where teams can see the current state of a project in real time.

When everyone is working from the same information, teams gain better visibility to help them make faster, more informed decisions throughout the project lifecycle. Just as importantly, it allows organizations to learn from one project and apply those insights to the next. Historically, much of that knowledge lived in spreadsheets, emails, and individual team members’ experience, making it difficult to carry forward. By capturing and structuring supply chain data across projects, Krane helps teams make better procurement and planning decisions over time.

Krane integrates with platforms already used across the industry rather than attempting to replace them. Do you think the future of construction AI will be built around the augmentation of existing software ecosystems or entirely new operating systems for construction?

In the near term, augmentation is essential.

Construction companies have invested heavily in platforms like Autodesk, Procore, Microsoft Project, CMiC, SharePoint, Trimble Viewpoint, and Oracle Primavera P6. Asking them to replace those systems isn’t realistic, and frankly, it’s not necessary.

Our approach is to integrate with the tools teams already use and create a connected layer across them. That allows customers to continue working within familiar workflows while benefiting from AI-driven procurement, materials management, and supply chain coordination.

Over time, I think we’ll see more AI-native systems emerge. But the companies that succeed will be the ones that fit naturally into the broader construction technology ecosystem rather than forcing customers to start from scratch.

The broader AI industry is moving rapidly toward agentic workflows. What lessons from construction could other industries learn about deploying AI agents in environments where real-world execution and logistics matter just as much as digital workflows?

One lesson is that AI becomes far more valuable when it’s connected to operational data.

Procurement OS works because it isn’t built on theoretical models of how construction supply chains should function. It’s built on the actual movement of materials across active projects. Every decision made in Krane creates data that helps improve future project execution.

We’ve seen that resonate with customers ranging from healthcare systems to some of the largest contractors in the industry, including Boldt, HITT Contracting, UCSF Health, and Juneau Construction. They’re not looking for AI that simply generates insights; they want systems that help teams make better decisions and execute work more efficiently.

Another lesson is that execution matters more than insights. Identifying a problem is useful, but solving it is where value gets created. In industries where physical operations and logistics matter, AI needs to help organizations take action.

Looking ahead five years, what does a fully AI-coordinated construction project look like? Which decisions will still belong to humans, and which parts of project execution do you expect intelligent systems to manage autonomously?

I think we’ll see AI become deeply embedded across the construction supply chain.

Tasks that are highly repetitive, administrative, and consume project teams will become increasingly automated.

In many ways, we’re already moving in that direction with Procurement OS and our AI construction crew. The vision is a system that learns from every project and helps teams make better decisions earlier in the project lifecycle.

That said, construction will always be a people business. Project strategy, stakeholder management, commercial negotiations, and major trade-off decisions will continue to require human judgment and experience. AI will handle more of the operational heavy lifting, while people focus on leadership, relationships, and decision-making.

Thank you for the great interview, readers who wish to learn more should visit Krane.