Lupa Technologies
AI-powered construction data intelligence platform that turns project documents into actionable insights for claims, disputes, and project optimization
lupa.technologyProduct Overview
What It Is
Lupa is a construction data intelligence platform. It analyzes millions of project documents across disconnected systems and turns them into searchable, source-linked insight for claims, disputes, project controls, and forensic analysis.
The Problem It Solves
Current Approach
Construction claims and project controls rely on manual review of millions of documents spread across more than 50 different platforms. Teams spend countless hours recreating project chronologies by manually reading through emails, RFIs, drawings, schedules, and reports stored in disconnected systems like Outlook, SharePoint, Procore, and various document control platforms.
How It Changes
Lupa automates the analysis of massive construction datasets, two to three million documents per project, connecting fragmented data silos and providing AI-driven insights in minutes rather than months. It enables data to speak to users through contextual search, automated reporting, and intelligent document analysis.
Area of Focus
The platform supports claims management and dispute resolution, provides live project controls and risk management, facilitates design change analysis and coordination, and enables forensic project analysis.
Key Features
Extract Reports and Data Intelligence
Analyzes a project's documents and returns an immediate Extract IQ Score showing data completeness and highlighting any gaps. It generates AI-driven narratives that explain what happened on a project, identifying key issues, critical timelines, responsible parties, and communication sentiment patterns. Every insight links back to the original source document for validation.
Weekly Disruption Analysis
Breaks an entire project into weekly data chunks, automatically identifying and categorizing problems such as design conflicts, material delays, and quality issues, with real-time alerts on live projects. This enables precise week-by-week reconstruction of events, which is especially valuable when preparing and supporting claims.
Intelligent Chat Interface
Users chat naturally with their project data. The assistant is aware of all project documents and communications and always provides direct source links for verification. It is especially useful for preparing for negotiations and meetings by analyzing patterns in correspondence.
Contextual Search and Bulk Tagging
Goes beyond keyword search by understanding context, so a search for material defects finds all the relevant issues. It automatically tags documents against major contract provisions including FIDIC, AIA and NEC, processes handwritten notes and scanned documents, and supports 115 languages, which makes it suited to international projects.
Drawing Analytics and Change Detection
Automatically compares thousands of drawings and creates visual overlays showing precisely what changed between revisions. It identifies unclouded design changes, which account for roughly 30 percent of typical changes, and generates narratives explaining the real impact of each modification.
Advanced Analytics
Schedule analytics compatible with Primavera P6, Asta Powerproject, and Microsoft Project, plus photographic analysis capable of detecting construction defects. The platform also produces deep research reports and applies retrieval-augmented generation tailored to different use cases.
Used By
Target Customer Types
Turner Construction
$14.9B revenue, 10,000 employees
“”
—
Skanska USA
$8.1B revenue, 9,500 employees
“”
—
Operating Locations
Primary markets: North America (US and Canada), the United Kingdom, and Australia and New Zealand. Expanding markets: the European Union, the Middle East, and Asia-Pacific.
Implementation & Onboarding
Learning curve: roughly two weeks to master all features. Processing time: about five minutes per gigabyte of data. Upload is as fast as the connection allows, and large datasets of three to four terabytes process in three to four days. Things to know: AI outputs are only as reliable as the data provided. Complex schedule analytics depend on the quality of the imported programmes, and deep-research models may miss context if documents are poorly scanned. The platform accelerates expert workflows but does not replace professional judgment.
Integrations
ROI & Value
Document review efficiency
More than 80 percent reduction in document discovery and review time.
Processing speed
One million documents processed in under two hours.
Cost savings
Hundreds of thousands of dollars saved per dispute.
Issue detection
Early identification prevents costly escalation.
Team focus
More than 80 percent shift from low-value tasks to high-value analysis.
Report generation
Cuts production time from weeks to hours.
Founding Team
Vladimir Milovanovic
CEO and Co-Founder
Vladimir brings over 15 years in construction technology and project management. He is a former McKinsey consultant with deep expertise in construction analytics.
LinkedIn
Milan Jankovic
CTO and Co-Founder
Milan is an AI and machine learning specialist with over 12 years in data science. He previously led engineering teams at major technology companies focused on document analysis.
LinkedIn
Đorđe Nedeljković
Chief Technical Officer
Đorđe has deep expertise in knowledge management, text mining, machine learning, and data analysis. He holds a PhD in Information Technologies in Civil Engineering and has seen first-hand the limitations of traditional data management tools in the industry.
LinkedInFunding
Total Raised
$1.8M
Stage
Seed
Investors