Blog / Information Management
Newforma does not replace Microsoft Teams. It does not replace Outlook.It does not replace point solutions. It does not replace Procore, Revit, ProjectWise, Bluebeam, Autodesk, or the systems firms already depend on.
That is not a flaw in the product.
That is the point.
AEC firms will continue to use email, drawings, models, documents, workflows, and chat. The idea that one platform can replace every communication channel, project system, and collaboration tool was wrong a decade ago, and it is still wrong today.
What the industry needs is not another destination.
It needs a layer that works across the tools teams already use.
A layer that captures the information flowing through them.
Ties it back to the project record.
Preserves the context.
Structures the decision trail.
Makes it timestamped, searchable, retrievable, and defensible.
That is Project Information Management.
It is the difference between scattered files and the full story behind every decision, change, handoff, approval, markup, RFI, submittal, and conversation.
When litigation arrives, when regulations require a decision log, when a client challenges a change, or when a project team needs to reconstruct what happened months later, firms with connected project information have the receipts.
Firms without it risk what Carl V calls “knowledge bankruptcy”
Not because their tools failed.
Because their project record did.
Edited Podcast Transcript
Participants
– Randall Stevens — Co-host, Confluence Podcast
– Evan Troxel — Co-host, Confluence Podcast
– Carl Veillette — Guest, Newforma
Editorial note: This transcript has been edited and condensed from YouTube’s automated captions for clarity, readability, and web publication. Repetition, filler, dated event promotion, and YouTube interface text have been removed. The meaning of the substantive discussion has been preserved. Host contributions are grouped because the automated captions did not identify individual speakers.
Introduction
Hosts:
Welcome to another Confluence podcast. Randall Stevens and Evan Troxel welcome Carl Veillette of Newforma back to the show, nearly a year after his previous appearance. The conversation picks up where the earlier discussion left off: the growing complexity of the AECO technology ecosystem, the rapid development of AI, and the challenge of connecting project information across an expanding number of applications.
Hosts:
The hosts ask Carl to begin with two ideas he has discussed frequently over the past year: information sprawl and the golden thread. They note that project teams continue to add tools and communication channels, making it increasingly difficult to preserve context and understand how decisions were made.
The Golden Thread and Tool Sprawl
Carl Veillette:
The technology stack is growing quickly. A few years ago, Newforma asked firms how many tools they typically used on a project, and the average response was about 15. In the Confluence technology stack survey, that number had risen to 84 applications across an organization.
Carl Veillette:
That growth is not necessarily negative. Teams are selecting best-in-class tools for specific tasks, which gives the industry more choice and allows people to work in the environments that suit them. The difficulty is that every additional tool can create another storage location, another data silo, and another governance challenge. Information becomes harder to connect, search, analyze, and preserve.
Carl Veillette:
The golden thread is especially important in the United Kingdom because regulation has increased the need for a reliable record of project decisions. Carl describes the golden thread as the ability to look back 10 or 15 years after a project is complete and reconstruct the audit trail: what happened, when it happened, who communicated with whom, and what the project context was at the time.
Carl Veillette:
For Carl, the golden thread is not the same as a single source of truth. He does not believe one system can become the only authoritative source for every kind of project information. Projects use multiple sources of truth. The practical goal is to connect those sources and create a single view of the information.
Carl Veillette:
A common data environment is therefore not necessarily one destination. In practice, it is a network of interconnected tools exchanging information. As the technology stack expands, the challenge is to connect those systems without forcing teams to abandon the tools they rely on.
Why Newforma Connects Rather Than Replaces
Carl Veillette:
Technology adoption is fundamentally a change-management problem. People have established ways of working, and they do not want to change tools simply because another platform has been introduced. Newforma’s approach is not to become the centre of every workflow. It is to provide an information layer that works across the best-in-class tools already used by project teams.
Carl Veillette:
Microsoft Teams is a good example. Teams has become an important communication channel and is unlikely to be displaced. Email is another example. For years, the industry expected instant messaging and collaboration platforms to replace email, but that has not happened. Email volume is increasing as projects become more complex, teams grow, and regulatory requirements expand.
Carl Veillette:
Project information now flows through Teams messages, meeting minutes captured with AI, email, design applications, construction platforms such as Procore, and many other systems. Trying to replace every source would create more disruption. Newforma instead connects those sources, collects the relevant information, and ties it back to the project context.
Carl Veillette:
The objective is to preserve the complete project record from the earliest stage through delivery and beyond. When a dispute or litigation occurs, the firm should be able to locate the evidence, communications, documents, and decision history needed to understand what happened and respond confidently.
Hosts:
The hosts refer to an example from Carl’s previous appearance in which a simple thumbs-up reaction in a message was later interpreted as tacit approval. They ask whether AI could help teams identify ambiguous communications proactively, rather than only helping people search the historical record after a problem occurs.
AI, Search and Information at Scale
Carl Veillette:
AI can support several parts of the information lifecycle. Newforma’s first AI implementation focused on collecting information. The system scans a user’s inbox, identifies project-related email, and associates that communication with the appropriate project record.
Carl Veillette:
Collection is only one use case. The volume of project data has grown so quickly that teams are reaching a practical human limit. Applications make it easier to create files, models, messages, reports, and other records, but more data does not automatically produce greater productivity. Individual tasks may become more efficient while the overall project suffers from information overload.
Carl Veillette:
AI and large language models can help teams search, summarize, and interact with project information at a scale that would otherwise require many people. A project team could ask questions across millions of data points and receive a summary of the current state, the relevant history, or the documents associated with a decision.
Carl Veillette:
Search may not sound as exciting as some AI applications, but it can have a significant operational impact. Better search supports informed decisions because people can find the relevant information when they need it, rather than manually reviewing large project archives.
Carl Veillette:
Code compliance is one potential example. Experienced professionals are retiring, and firms are struggling to preserve the knowledge built through years of practice. New graduates and junior employees may not yet understand every building-code requirement or local regulation. AI could act as a safety net during document review by identifying a potentially missing requirement before information is issued for construction.
Carl Veillette:
The purpose would not be to replace professional judgment. It would be to support onboarding, quality control, and knowledge transfer. AI could flag an omission, prompt a reviewer to investigate, and reduce the likelihood that a preventable issue becomes a dispute later in the project.
Agentic AI and Institutional Knowledge
Hosts:
The hosts ask how Newforma could move from a reactive historical use case to more proactive assistance during project delivery.
Carl Veillette:
Newforma has information associated with more than 20 million projects delivered using its systems. Carl sees that scale as a foundation for agentic workflows. Future agents could support activities such as code review, RFI responses, knowledge discovery, and lessons learned.
Carl Veillette:
He describes a possible knowledge agent. Imagine a team designing a green roof. The firm may have completed a similar system on another project, but the previous team may have called it a sedum roof. A knowledge agent could recognize the semantic relationship, find the earlier project, surface the related documents, and show whether that design resulted in RFIs, change orders, or construction issues.
Carl Veillette:
That context could help the current team avoid repeating a past mistake. The value comes not only from possessing the documents, but from understanding the relationships between the design, the RFI, the contract change, and the eventual outcome. Without the golden thread, it is difficult to create reliable predictions or recommendations.
Hosts:
The conversation turns to building codes and jurisdiction-specific requirements. The hosts note that code information can differ by location and edition, so a responsible system must use the correct source and avoid presenting outdated requirements as authoritative.
Carl Veillette:
Carl agrees and describes a potential partnership model. Newforma manages processes such as issuing construction documents, while a specialist platform such as UpCodes maintains the code library. An integration could use the project address and jurisdiction to identify the applicable requirements and then support analysis of the relevant construction documents.
Carl Veillette:
The goal would be to combine workflow context with trusted specialist data. Newforma does not need to become the owner of every code library. It can connect the process, the project information, and the appropriate external source.
Onboarding, Proposals and Lessons Learned
Hosts:
The hosts observe that users are often the weakest link in archiving. Every project is a new team with new constraints, and people are not consistently good at organizing information for their future colleagues. At the end of a project, firms can effectively declare ‘knowledge bankruptcy’ and begin the next job without preserving what the previous team learned.
Carl Veillette:
Carl gives the example of a new intern asked to support a proposal. The intern may need to find previous hospital projects above a certain value, but the relevant information could be distributed across a CRM, SharePoint, local file servers, and other systems.
Carl Veillette:
Because Newforma connects to those sources, an assistant could allow the user to ask for all relevant proposals regardless of where the files live. It could then help assemble a new proposal using an approved template, project examples, and staff resumes.
Carl Veillette:
The underlying information already exists; the problem is that it lives in separate silos and is difficult to discover. AI can support onboarding, training, quality control, and the transfer of collective knowledge built by the firm over many years.
Carl Veillette:
The same principle could apply during project work. While a team member reviews a document, an assistant might indicate that a similar detail led to litigation on a previous project, explain the cost or outcome, and surface the associated record. The current team may be completely different from the team that experienced the original problem, so the system must carry that knowledge forward.
Data Quality, Timestamps and Trusted AI
Hosts:
The hosts ask whether organizations can simply provide every available file to an AI system. More data may seem better, but project archives contain superseded documents, informal communications, incomplete records, and information that may no longer be correct.
Carl Veillette:
Carl’s answer is straightforward: garbage in, garbage out. Sending a large volume of unstructured information to an AI system without context is risky. Consider a contractor searching construction drawings on site. If the system cannot distinguish between revisions, it may answer a question using a superseded drawing and contribute to an incorrect concrete pour or installation.
Carl Veillette:
The structure behind the information is therefore critical. Newforma timestamps project records so the system can understand the sequence of events and create a conceptual timeline. It can distinguish between an early email, a later formal document, and a superseding revision.
Carl Veillette:
Without that structure, an AI system may use an unofficial email instead of the latest construction document or treat conflicting information as equally valid. A raw project archive placed into a general chatbot would likely produce unreliable answers because the system would not understand status, authority, revision, or project context.
Carl Veillette:
Reliable AI requires a backend system that collects, classifies, timestamps, and relates the information before the model is asked to interpret it. The information-management foundation is what makes the resulting search, summary, or recommendation trustworthy.
Hosts:
The hosts acknowledge that AI capabilities are improving rapidly and may eventually become better at judging relevance. Even so, the current challenge remains: firms need contextual metadata and a reliable record so that any human or AI system can determine which information should be used.
Why Data Matters More Than the AI Layer
Carl Veillette:
Carl challenges the idea that AI will automatically displace major enterprise platforms. AI is only as useful as the data available to it. Enterprise systems remain valuable because they contain the structured business and project information needed to train, ground, and operate AI capabilities.
Carl Veillette:
AI is a mechanism for processing and interacting with data. The long-term value lies in the information, its quality, and the relationships captured around it. Organizations may eventually create many different AI experiences, but without reliable data those experiences will have little substance.
Hosts:
The discussion then returns to the Confluence technology stack survey. More than 400 applications were included and grouped into broad categories. The hosts ask where Newforma belongs because it overlaps several areas without performing the same work as the applications it connects.
Newforma’s Role in the AECO Technology Stack
Carl Veillette:
Newforma occupies an unusual position because it participates in several workflows without trying to replace the specialist application where the work is performed. In communication, Newforma connects with Microsoft Teams and Outlook but is not itself a messaging or email application.
Carl Veillette:
In enterprise resource planning, Newforma can connect with ERP systems and reuse information so that teams do not have to create the same project data twice. In construction administration, it supports designer-side workflows and integrates with contractor platforms such as Procore rather than attempting to become a contractor’s construction-management system.
Carl Veillette:
In BIM coordination, Newforma is not a clash-detection or authoring platform. It connects coordination information between tools such as Navisworks and design applications, tracks issues, preserves accountability, and creates the workflow history and audit trail.
Carl Veillette:
Project Information Management is the broad category Newforma has used for approximately 20 years. The purpose is to connect the dots between communication, documents, workflows, and project systems so that firms can maintain the golden thread.
Hosts:
The hosts point out that customers may know Newforma for different parts of this value proposition. One user may think primarily of search, another of email management, another of construction administration, and another of BIM issue tracking. A brand is often defined by the way customers use and describe it, not only by the category selected by the vendor.
Hosts:
The survey also showed that firms rely on many storage and content-management locations. A typical organization may use SharePoint, Autodesk Construction Cloud, local servers, cloud storage, and other repositories. The end user must know which system contains the information, which adds friction and reinforces data sprawl.
Connecting Storage Without Becoming the Storage Platform
Carl Veillette:
Newforma does not need to host every file. It connects to the sources already selected by the organization, including SharePoint, Autodesk Construction Cloud, local file servers, ProjectWise, and other storage environments.
Carl Veillette:
The objective is to provide a unified view, support information sharing, apply retention and document-control processes, and help firms meet standards such as ISO 19650. The project may contain infrastructure files in ProjectWise and Revit files in Autodesk Construction Cloud. Newforma connects those locations so the organization can maintain the audit trail without moving everything into another repository.
Carl Veillette:
This neutral position is important. The technology stack will continue to vary by client, project, discipline, and region. An information-management layer must work across that diversity instead of requiring every participant to adopt the same storage system.
Governance, Retention and Regulation
Hosts:
The hosts ask when governance became a major concern in the industry.
Carl Veillette:
Carl believes the issue has become much more visible over the last five years. Cloud adoption, AI, and the proliferation of applications have forced CIOs and IT teams to understand where information lives, how it flows, and whether the organization will be able to use it in the future.
Carl Veillette:
Regulation is also increasing the pressure. The Building Safety Act in the United Kingdom, for example, requires stronger documentation and decision records. Contracts have long included retention requirements, but complying with those obligations becomes more difficult when data is spread across a growing number of platforms.
Carl Veillette:
Retention is not always about keeping information forever. Some organizations want the record available so they can defend themselves in a dispute. Others deliberately delete data after the contractual retention period because retaining it creates cost and may create additional discovery obligations during litigation.
Carl Veillette:
The organization must therefore understand what information it has, which rules apply, and when the record should be retained or deleted. Governance needs to sit above the various systems and support consistent policy across the project portfolio.
The Content Router and the Cost of Disconnected Systems
Hosts:
The hosts describe project information as a capital resource. Files and records need contextual wrappers that explain what they are, where they belong, how long they remain useful, and where they should go. Traditional software asks a person to choose a location every time they use File Open or File Save, placing responsibility for information routing on the user.
Hosts:
A future information system could behave more like a network router. Instead of asking people to make every filing decision manually, it would use rules and metadata to direct information to the correct project context, apply governance, and preserve the information needed later.
Hosts:
This is difficult because firms use many applications, each producing information in a different format. Project managers are often assigned responsibility for archiving, but teams may not receive the time or structure needed to complete what is effectively a major closeout task.
Carl Veillette:
Carl says Newforma frequently hears this concern from firms after litigation increases or a case is lost because the organization cannot produce the necessary evidence. Those firms begin formalizing processes, pursuing standards such as ISO 9001, and searching for an information backbone that can structure collection and archiving.
Fixing the Seams Between Tools
Hosts:
Every application may perform its own task well, but the workflow often breaks at the seam between two systems. Firms then purchase a third tool or create a manual process simply to move information between the first two. The cost appears as extra logins, duplicate data entry, uncertainty about file versions, and missing audit trails.
Carl Veillette:
APIs and native file formats are both part of the challenge. Proprietary formats make information harder to consume after the original application changes or disappears. Newforma has long invested in indexing industry formats such as CAD and Revit files so the information can be searched and understood without requiring the user to open every source application.
Carl Veillette:
The goal is to free the data from the individual file and make the relevant content discoverable. This is why Newforma customers often describe the product as the Google search of construction projects. Search is one of the clearest ways users experience the value of connected project information.
Hosts:
The hosts conclude that future technology surveys may need to allow platforms to belong to multiple categories. Collaboration, content management, information management, governance, and search increasingly overlap, particularly for systems designed to connect rather than replace other applications.
AI-Written Code and Closing Thoughts
Hosts:
The final discussion considers AI-written software and ‘vibe coding.’ If individuals throughout an organization can create small applications, software sprawl and information sprawl may accelerate. Firms may gain useful tools quickly, but they will also have more code, more data sources, and more governance questions.
Carl Veillette:
Newforma already uses AI in software development. The speed advantage is significant, so avoiding AI-generated code is not a realistic option. At the same time, the company has increased its use of security-scanning tools because AI may generate and validate code with fewer people directly involved in every step.
Carl Veillette:
The same principle applies to architecture, engineering, and construction firms. Organizations that refuse to explore AI may struggle to remain productive and competitive. The responsible approach is to adopt the technology while strengthening data quality, security, governance, and human oversight.
Hosts:
The episode closes with the expectation that these questions will continue to evolve quickly. AI may change how software is created and how project information is used, but the need for connected, structured, traceable information will become more important—not less.
About the Author
