The Challenge
A mid-size architecture practice faced a substantial challenge in efficiently handling the vast amounts of data generated through their Revit models. These models were rich in detail, containing vital information crucial for making informed decisions throughout a project's lifecycle. However, the practice struggled with integrating this data into a coherent and easily accessible format that could deliver real-time insights.
The client needed a solution that would transform their Revit model data into a dynamic, live project intelligence portal. This portal was required to streamline data management, provide on-the-fly updates, and enhance collaboration across different project stakeholders.
The Solution
Adyantrix delivered a robust BIM API integration, seamlessly connecting Revit model data to a custom-built live project intelligence portal. Our team meticulously developed a set of APIs that extracted essential data from Revit models and channelled it into a cloud-based intelligence platform.
This integration was facilitated through a user-friendly dashboard designed to cater to non-technical team members. Key data points from Revit were mapped to the portal, offering filtered access and visualization tools that helped decipher complex datasets into actionable insights.
Collaborating closely with the client's design and IT teams, Adyantrix streamlined API endpoints to ensure fast data updates, minimising latency issues. Additionally, we incorporated robust security protocols to safeguard sensitive information against potential data breaches.
Key Results
The BIM API integration yielded impressive results, transforming the architecture practice's data management and operational efficiency.
- Improved Data Accuracy: By connecting Revit to the live intelligence portal, the practice witnessed a 35% reduction in data discrepancies, ensuring that all stakeholders had access to the most accurate and up-to-date information.
- Enhanced Collaboration: The portal facilitated better communication and coordination among teams, leading to a 40% improvement in workflow efficiency and faster decision-making processes.
- Time Savings: Automated data updates reduced manual handling, saving over 200 hours annually that were previously invested in manual data entry and corrections.
- Revenue Growth: Enhanced project insights and streamlined operations contributed to a 15% increase in project ROI, as stakeholders could make more informed decisions, reducing errors and rework.
The successful implementation of a dynamic BIM API integration has equipped the architecture practice with the tools necessary for enhanced project management and long-term strategic growth. As a result, they continue to embrace digital transformation, realising the full potential of their Revit investments.
Technical Approach
The integration was built around three distinct API layers: a Revit add-in that ran server-side within the Autodesk Desktop Connector environment, an intermediary data transformation service, and a REST API consumed by the portal front-end. This three-tier architecture was chosen because Revit's API is synchronous and document-bound — it cannot be called directly from a web application — requiring an intermediary layer to bridge the gap between Revit's desktop execution context and the cloud-hosted portal.
Key technical components included:
- Revit C# add-in using the Revit API 2024 SDK: The add-in hooked into Revit's
DocumentSavedandDocumentSynchronizedWithCentralevents, triggering an automated data extraction run whenever a model was saved or synced to the central file. Extracted data included element counts by category, parameter values for tracked families, sheet issue status, revision history, and Navisworks clash counts pulled from embedded clash test results. - Autodesk Platform Services (APS) for cloud model access: For models hosted in Autodesk Construction Cloud, we used the APS Model Derivative API to extract structured data from uploaded IFC and Revit files without requiring a local Revit installation, enabling the portal to process models uploaded by any team member without triggering the desktop add-in.
- Azure Functions for the transformation service: Lightweight serverless functions handled the normalisation and enrichment of raw Revit data before writing to the portal database, including currency conversion for cost data, ISO 19650 compliance validation, and deduplication of elements reported by both the desktop add-in and the APS pipeline.
- PostgreSQL with JSONB columns for flexible schema: Because different project types required different data points, we used JSONB columns to store project-specific parameter sets alongside the core structured schema, allowing the portal to accommodate new data requirements without schema migrations.
- React front-end with D3.js for data visualisation: The portal presented model progress, clash status, and issue resolution trends through interactive charts. Role-based access control allowed clients to see a read-only summary view while internal project managers accessed the full data set.
Implementation Highlights
The initial challenge was the heterogeneity of the practice's Revit models: different projects had been set up by different teams over several years, using inconsistent parameter naming conventions and family naming standards. The extraction add-in had to accommodate this variability rather than assume a uniform data structure. We built a configuration layer — a per-project mapping table maintained in the portal's admin interface — that allowed the portal administrator to define which Revit parameters corresponded to which portal data fields for each project. This added initial setup overhead per project but eliminated the alternative, which would have been to standardise all existing Revit templates simultaneously — a multi-month exercise that would have disrupted live project work.
One unexpected technical challenge arose with the Revit DocumentSynchronizedWithCentral event: in workshared models with many concurrent users, this event could fire multiple times within a short window as different users synced, potentially triggering redundant extraction runs and writing conflicting intermediate states to the portal. We resolved this with an idempotent extraction pattern: each extraction run was assigned a hash of the central model's current sync timestamp, and duplicate hashes were discarded by the transformation service before database writes.
The APS integration required navigating Autodesk's authentication model, which uses OAuth 2.0 with both two-legged (service-to-service) and three-legged (user-delegated) flows. We used two-legged authentication for automated extraction runs and three-legged for user-initiated actions such as requesting a fresh clash detection report — a distinction that satisfied both Autodesk's API usage policies and the practice's information security requirements.
The portal's client-facing summary view was designed through two rounds of co-design workshops with the practice's most active client contacts. The key insight from those workshops was that clients wanted to see two things above all else: the percentage of the model at the agreed LOD milestone, and the number of outstanding issues — expressed as a trend line, not just a current count. Both required specific data extractions and aggregations that were not initially scoped, but adding them based on client feedback before launch resulted in a portal that clients described as genuinely useful rather than merely informative.
Measurable Outcomes
The 200-hour annual saving in manual data entry and corrections was measured against a baseline established during the discovery phase. Before the integration, the practice's BIM manager spent approximately four hours per project per week manually compiling model progress reports from Revit, Navisworks, and the project management tool — a task that the integration automated entirely. Across the eight active projects in scope at launch, this represented a recurrent saving of 32 hours per week of senior BIM manager time.
The 35% reduction in data discrepancies was most visible in the reduction of clashes between the model data reported in project meetings and the model data visible to the site team. Previously, the lag between model updates and reporting meant that clash counts in project meeting reports were typically two to three weeks out of date. With the integration, the portal reflected the current clash status within two hours of any model sync — giving all parties a shared, current view of the model's coordination status.
The 15% increase in project ROI was attributed primarily to faster design decision cycles. When a design query arose — for example, whether a particular floor's MEP coordination had been signed off — the answer was available immediately from the portal rather than requiring a manual search of the project archive or an email to the BIM manager. Faster decision resolution reduced the knock-on programme delays that such queries had previously caused, with the largest single saving being a three-day reduction in the design freeze delay on the practice's largest active project.
Lessons Learned
The most important lesson was the value of event-driven extraction over scheduled polling. Our initial design used a 30-minute polling job to check for model updates. In practice, this introduced up to 30 minutes of staleness in the portal data, which was noticeable during design review sessions when model changes were made in real time. Switching to an event-driven approach — triggered by the Revit save and sync events — reduced the average lag to under five minutes and eliminated the "why doesn't the portal show the change I just made" frustration that had been undermining user confidence.
We also learned that the add-in installation and update process deserves as much attention as the add-in functionality. Revit add-ins are installed per user per machine, and a practice with 200 users across three offices has a complex software deployment environment. We built an auto-update mechanism into the add-in that checked a version endpoint at launch and prompted users to download the latest version if their add-in was out of date. This reduced the version fragmentation problem that had made our early testing inconsistent.
Why This Approach Worked
The integration succeeded because it met the data where it lived rather than requiring the practice to change how they worked. Architects and engineers were not asked to adopt a new tool, enter data into a separate system, or change their Revit workflows. The add-in and APS pipeline operated invisibly within the existing save and sync events that the team was already performing. The only visible change for the design team was that the portal was automatically up to date after every sync — something that felt like a quality improvement rather than an additional task.
For the practice's clients, the portal provided a level of transparency that differentiated the practice from competitors still relying on weekly PDF status reports. Two new clients cited the portal's live visibility of model progress as a factor in their decision to appoint the practice — a commercial benefit that extended well beyond the operational efficiencies the integration was originally designed to deliver.
Speak with our Custom Software Development team at Adyantrix to find out how we can support your next project.
Work with Adyantrix
If you are looking to tackle a similar challenge, Adyantrix has the expertise to help across the full project lifecycle. Our custom software development practice covers tailored applications built to your exact workflows. Our data engineering practice covers pipeline design, streaming, and data infrastructure. Our business intelligence practice covers BI strategy, platform selection, and dashboard delivery. Our BIM automation practice covers Dynamo scripts, Revit plugins, and model quality automation. Get in touch to discuss your requirements — no commitment required.
