See how Willow can help you cut costs, save energy, and operate your buildings more efficiently and sustainably.
Up to 95% of the information generated and captured in documents during building design and construction can go unused in building operations, according to research. For built environments, documentation is essential but managing it can be overwhelming. Facility teams have to track architectural packages, O&M manuals, floor plans, and more.
All this information is often spread across systems such as BMS, CMMS, and standalone document repositories. Disjoint storage systems, unclear version control and inconsistent tagging practices can make it difficult to access information quickly. Maintaining a real-time view of assets, understanding where they are physically located versus spaces they serve, and which documents accurately reflect their current state remains a persistent challenge. Contextually locating relevant information buried in text or images, whether identifying a fault code, or checking for compliance can be time-consuming and frustrating.
Willow transforms this experience with AI-driven techniques. Below, we explore five key capabilities that reshape how documents are utilized through the lifecycle of building management.
Organizations share a common challenge when it comes to generating a digital representation of assets and spaces from documents and then maintaining a real-time view over time. Spread across BMS, CMMS, and construction documentation, asset information can drift through lifecycle events. It can be difficult to maintain continuity across design drawings, construction submittals and as-builts. As a result, asset truth degrades over time, forcing repeated reconciliation and undermining operational and capital decisions.
With Willow, onboarding is fast and easy. Sub-graph generation with AI automates what can otherwise be a labor-intensive process, requiring teams to manually review architectural drawings and TAB reports, extract asset information, and build out inventories by hand. A rich ontology describes models for space and asset types. Uploaded documents are analyzed to hydrate the Knowledge Graph with twins that uniquely represent physical assets, HVAC zones, locations, and their relationships.
As integrations with BMS, Occupancy and CMMS light up, asset and space identifiers from each system are associated with unified entities. This ensures that for each asset and space, a single twin in Willow continues to represent its physical counterpart. The Knowledge Graph comes to life as telemetry starts flowing, reflecting the systems and structure of the real environment. Over time, as assets are added or moved, new twins are created and relationships to spaces are updated to keep the graph up to date to match device lifecycle events in the real world.
As twins are created and subsequently kept up to date, it becomes easy to filter for assets of a specific model and capture counts for inventory. Docs are leveraged to accelerate onboarding and enable scale across global portfolios.
Certain classes of documents are most relevant in context of specific assets and spaces. For instance, teams may want to retain the original warranty documents for boilers and chillers. Troubleshooting guides and PM schedules and in O&M manuals apply to all AHUs from a specific manufacturer. Willow makes it easy to preserve these documents in the Knowledge graph with 1:1 or 1:many relationships with twins for assets and spaces.
In this example, the twin for an Air Handling Unit called AC-10-1 is related to documents for fault codes, a winter readiness checklist and an O&M manual provided by the manufacturer. Over time, teams can curate knowledge base documents and capture images during inspection and maintenance and relate to relevant asset twins. A fire safety plan might be linked to the spaces it governs. Compliance documents, evacuation diagrams, cleaning schedules, or lease information can all be associated with the assets and locations they apply to. These relationships ensure that when a user navigates to an asset or a room, relevant documents can be referenced in context.
A PDF viewer in Willow makes the contents of the uploaded document viewable. In addition, content can be queried via natural language in Willow Copilot.
Documents essential to operations like O&M manuals can run hundreds of pages long. Even with access to the correct file, locating the relevant table or procedure can be time-consuming. Willow solves this through Retrieval Augmented Generation (RAG) and natural language prompts. For technicians in the field, troubleshooting is simplified with queries such as “I got fault code 12. What should I do?” Prompts are processed by AI, retrieving the relevant information from a vectorized index of uploaded documents, starting with the ones directly related to the asset in context. Willow Copilot synthesizes responses into clear, grounded answers while providing citations to the original uploaded documents.
With Willow, critical and hard to find details buried deep in documents become accessible within seconds. Technicians can quickly get to information they need in context of repair and servicing, increasing operational efficiency and reducing equipment downtime.
A common challenge with large document sets is to quickly scan and determine what the contents are without opening each file. File names are often not descriptive enough. Willow’s solution is to leverage AI to automatically generate concise summaries for every document upon upload. The summary is then stored as metadata with the document. This makes it easy to determine at a glance in what the document contains, and for Willow Copilot to reference it in responses.
For long PDFs, AI summary can highlight key information like Manufacturer, part numbers, installation dates, warranty expiration and document creation details.
Beyond PDFs, customers can upload image files that traditionally are difficult to infer without opening and viewing. Photos taken during rounds, inspections or during technician visits may capture images of equipment, labels, or site conditions. When uploaded in Willow, AI summaries generate searchable text on the image details, with the file itself getting related to the relevant assets and spaces in the Knowledge Graph.
AI summaries of images also reveal insights that can influence compliance and risk management. In this example, a sprinkler head that got painted over during a renovation is flagged to be in violation of NFPA guidelines. As a result, teams can be alerted proactively.
Documents evolve as buildings change and systems are replaced. Without version control, uploading updated documents can result in duplicates and confusion about which file contains the most current information.
Willow handles versioning of documents seamlessly. When a new version is uploaded, the platform stores it alongside previous versions, preserving a historical record. Older documents remain accessible, allowing teams to track changes over time, while the latest version is actively used. This creates an environment where teams can be confident that they are working with the newest information.
Modern built environments depend on accurate, accessible information. Willow brings intelligence, structure, and automation to documentation by combining AI onboarding, graph relationships, RAG-powered Copilot search, AI summaries, and robust version control.
This approach integrates documents into a living, connected Knowledge Graph, related to assets and spaces. Teams are able find answers instantly and confidently make operational decisions. Buildings become simpler to manage, portfolios scale seamlessly, and organizations realize a new dimension of operational efficiency.