01
Identify
Connect the placed object to the asset and version it represents.
Collaborative Asset Intelligence
Asset intelligence
across tools and teams.
BRAID connects 3D assets to their versions, origins and project use. Teams share that context across the visualisation lifecycle. Planned AI capabilities build on this documented evidence to answer project questions and coordinate work.
Understand how they connect.Keep track of sources, variants, project use and relationships. Collaborate through the web, engines and creative tools.
The coordination problem
Project context is
scattered across systems.
Which version is in use? Where did it come from? Who else will a change affect?
Teams piece together answers from file paths, messages and individual memory. The work repeats whenever a project moves between tools or organisations.
BRAID gives each asset a persistent identity and shared record that connects to your existing tools and source systems.
Current capabilities
Web and Unreal Engine
share scene changes.
02
Coordinate
Keep placement changes aligned between the web application and Unreal Engine.
03
Retain
Keep a record of changes, decisions and project use.
The August 2026 Test/Showcase release demonstrated scene coordination between the web and Unreal Engine, alongside managed asset delivery. Private beta will test onboarding, reliability and scale with more customers.
Collaboration across the visualisation lifecycle
From the first scene
to the final publication.
Collaboration continues from pre-visualisation through production, review and distribution. Our roadmap carries asset identity, versions and decisions between these stages, so each team can use the work and knowledge already gathered.
Pre-visualisation
Acquire assets, explore creative ideas and agree on the project direction.
MeasureTime to first contribution
Production
Coordinate changes in the web and Unreal Engine, with Blender and Unity coming soon.
MeasureVersion alignment & coordination effort
Post-production
Review with context. Connect feedback and decisions to the work.
MeasureReview turnaround & rework
Distribution
Carry provenance, approvals and licensing context into publication and reuse.
MeasureContinuity between teams & reuse
Pilot teams will measure these outcomes as integrations extend support across the lifecycle.
CASE STUDY: Royal Commission for AlUla ↗
Coordinating 40+ vendors
and 10,000+ assets.
A shared workflow replaced manual file synchronisation, spreadsheet tracking and email approvals.
60%shorter review cycles
Version conflicts eliminated
Within the reported RCU deployment.
Minutes, rather than weeks
To resolve spatial drift in the reported workflow.
These results come from project debriefs, regular operational reviews and side-by-side analysis of project operations. We requested a trial of our methods and presented the results during delivery. RCU did not commission a formal evaluation. The results relate to this workflow.
Time and productivity
Give experts more time
for the work that matters.
Less time tracing changes and decisions means more time designing, reviewing and solving problems. The asset record keeps that knowledge available to the next person who needs it.
Time to productivity
Less preparation before
the first contribution.
Start with a lightweight project shell. Join the configured BRAID workflow while assets and instances localise in the background, without downloading the full asset library upfront.
Traditional
Full project localisation before work begins
BRAID
~1h engine setup, then connect and join Engine estimate; connection and startup time varies
Illustrative workflow. *20h+ reflects client-project experience. Shell size and engine time are planning estimates; <1s covers ideal payload transfer only. Segments show stages, not elapsed time. Access and plugins are preconfigured; required content and shaders may still need preparation.
Timing basis and project assumptions
Shared engine setup: approximately one hour
Both workflows allow one hour for Unreal Engine 5.7.4 on a Windows workstation with SSD: a 36 GB download budget at 100 Mbps takes around 48 minutes, with 12 minutes allowed for installation and verification. EGDATA’s Windows manifest provides the 36.06 GB full-download reference. Selected Launcher components, hardware and connection quality affect the actual time. Project-specific shader preparation follows separately.
Perforce localisation: 20h+ in client workflows
Client project experience shows wide variation across Perforce deployments. This comparison uses 20h+ for initial localisation of a 100 GB project, based on that delivery experience. Perforce documents how network latency, single-connection limits and decompression affect sync performance. Parallel sync, nearby proxies and existing caches can shorten the wait. Source upload limits and destination performance can introduce further bottlenecks. The ideal 100 Mbps transfer takes 2h 13m for 100 GB; full localisation includes the constraints of the deployment.
BRAID: a project shell under one megabyte
The model uses a 100–500 KB project shell in source control, with the asset library managed through BRAID. The shell contains the project descriptor, shared configuration, minimal startup content and any project source. At ideal 100 Mbps, its payload transfers in under a second. Connection, source-control file handling and editor startup determine the remaining joining time.
The size is a planning range for a lightweight shell. Compatible plugins, access and dependencies are preconfigured; additional startup content or custom code increases the initial payload. Unreal separates project source and content from generated build data and caches. BRAID then localises the assets and instances required by the work.
Readiness for individual tasks depends on the required content, shaders and plugin compatibility. Shared derived-data caches can reduce preparation in either workflow. P4 Cloud storage and egress are sized for the project, its history and team usage. The comparison centres on how much content must be local before a team can contribute.
Beyond the first session
Reduce coordination overhead.
Contain project overruns.
35%of time spent on non-productive work
Finding project information, resolving conflicts and correcting mistakes. The 2018 Autodesk/FMI survey shows the scale of the coordination problem.
Read the industry research ↗60%shorter review cycles in our trial
The confirmed RCU workflow result. Faster decisions can help prevent repeated reviews and blocked tasks from building into wider delays.
Measured review-cycle improvement. Overall project cost and schedule savings require separate validation.
Cost and schedule impact
What a 60% reduction in overruns would mean
McKinsey’s completed megaproject sample averaged 37% over budget and 53% over schedule. Applying a 60% reduction to those overruns illustrates the scale of the opportunity.
Illustrative project: US$3.8bn budget · 1,000-day programme
22.2%of original budget avoided in this scenario
37% cost overrun14.8% remaining
US$843.6m modelled saving
against US$1.406bn projected cost overrun.
US$562.4m of overrun remains in this scenario.
31.8%of original duration avoided in this scenario
53% schedule overrun21.2% remaining
318 days avoided
from 530 days of projected delay.
The programme moves from 1,530 to 1,212 days, leaving 212 days of delay.
The 2017 report identified more than 3,600 announced megaprojects with an average budget of US$3.8bn, and projected US$5 trillion in losses if historical performance continued.
Illustrative scenario, not measured whole-project BRAID savings. It assumes a 60% reduction across the full benchmark overrun. Actual impact depends on which causes the workflow can address. Cost and time percentages use different baselines and must not be added together. McKinsey, 2017: megaproject research ↗
Benchmarks and calculation details
The 35% figure describes respondents’ time spent on non-productive activities in the 2018 Autodesk/FMI survey of nearly 600 construction professionals. It is a measure of industry friction, not a claim that BRAID recovers 35% of every team’s capacity or a specific measure of software technical debt.
The dollar illustration uses the report’s US$3.8bn average budget for announced megaprojects. The 1,000-day programme is an illustrative duration, not a report average. Combining this budget scale with the completed-project overrun rates is a scenario, not an observed project result. Cost: US$3.8bn × 37% = US$1.406bn projected overrun; 60% avoided = US$843.6m, leaving US$562.4m. Duration: 1,000 days × 53% = 530 days of delay; 60% avoided = 318 days, leaving 212. No separate dollar value is assigned to those days, which could double-count the cost benefit.
The cost and schedule averages describe completed projects within McKinsey’s dataset of more than 500 resource and infrastructure projects above US$1 billion. These historical megaproject benchmarks are not averages for every construction project. See The art of project leadership (2017), pages 10 and 11. Sector, scale and the original estimate affect the comparison.
Scenario arithmetic: 37% × 60% = 22.2% of original budget avoided, leaving 14.8% overrun. 53% × 60% = 31.8% of original duration avoided, leaving 21.2% delay. Against the forecasts including overruns, the reductions are 22.2 ÷ 137 = 16.2% of cost and 31.8 ÷ 153 = 20.8% of duration. The 60% trial result concerns review-cycle time. Applying it to full project overruns is a hypothesis to test, not a demonstrated causal relationship.
If only half the overrun is addressable and that part falls by 60%, the avoided amounts become 11.1% of original budget and 15.9% of original duration. Financing, materials, scope changes and site conditions are examples of factors that may lie outside BRAID’s influence.
The earlier ~500-hour illustration remains valid as a small coordination example: 20 people × half an hour saved per week × 48 weeks = 480 hours. It does not estimate the full value of shared asset records or the downstream cost of delay. It is separate from the overrun scenario and should not be added to it without checking for double counting.
Initial market and expansion
Starting in visualisation.
Built for the wider 3D sector.
Architecture & the built world
Our initial focus: visualisation teams coordinating complex assets across project partners.
Industrial visualisation & digital twins
Keep asset history available for training, simulation and ongoing operations.
Film, VFX & real-time experiences
Track creative assets as they move between studios, tools and productions.
The roadmap covers asset acquisition and creative development, production, review and feedback, and licensing for distribution. Unreal Engine is supported today, with Blender and Unity coming soon. The web connects collaborators throughout.
The tools you know.
The commercial model
Per-seat subscriptions.
Free access for collaborators.
An agreed setup package covers integration and training.
Introductory subscription: US$25 to US$30 per user per month. List price: US$99 per user per month from month 12.
Storage, compute and database usage are charged separately. Limited free access lets clients, reviewers and partners join the workflow. Each invitation can introduce BRAID to another team, supporting growth through collaboration.
Request the commercial overview for pricing, implementation and unit economics ↗
The roadmap
Connected work.
AI built on project evidence.
Today
Asset delivery and scene coordination
The web and Unreal Engine share scene changes and asset delivery, establishing the project record that future intelligence will draw on.
Next
Private beta and wider tool support
Work with 5 to 10 pilot customers to measure onboarding, coordination and review times, alongside deployment effort and paid adoption. Extend production workflows to Blender and Unity.
Beyond
AI query and orchestration
Use retrieval-augmented generation (RAG), reranking and documented relationships to answer complex questions, assess changes and coordinate work across tools.
Pilot results and technical readiness will guide the release sequence.
Every answer has a project record behind it
Assets, versions, instances, placement and provenance connect to libraries, metadata, taxonomy, audit trails and change logs. The planned AI will retrieve records within the user’s permissions, trace relationships and cite the evidence behind each inference. Missing or conflicting information stays visible.
A question for the fully developed platform
“Find a public gathering space with trees, good sunlight after 3 pm, nearby amenities and a playground, within a five-minute walk of food and drink.”
One query across a connected 3D world
The planned AI would connect georeferenced spaces, placed trees, public-access records, amenities and walkable routes with sun and shadow analysis for the selected date. It could return candidate spaces, explain how each meets the brief and link to the underlying evidence.
Group size, accessibility and the date would define the search. Missing access records, incomplete routes or uncertain analysis would remain visible. Today, this kind of question typically requires manual work across several datasets and tools.
Richer intelligence, simpler access
Over time, BRAID’s context and orchestration would become accessible through skills and plugins for tools such as Codex, Claude and Gemini. Creators could choose work to publish in public libraries, with BRAID coordinating discovery, asset delivery, licensing and usage payments. Evidence, rights and history would stay connected as work is reused.
Why an independent platform?
Collaboration should work
across the tools you choose.
Visualisation platforms often improve collaboration within their own products while making it difficult to work across other tools. BRAID takes an independent approach, connecting asset records across those boundaries.
As teams connect more projects, they build a record of asset history and use. Its connected metadata, provenance and change history provide a documented basis for AI queries and inference across tools.
The execution partner
Experience at scale.
A team ready to deliver.
Neon Light combines architecture, visualisation and systems design experience from complex international projects.
That delivery experience informs how BRAID handles project coordination, technical requirements and the needs of the teams using it.
An experienced team is ready to mobilise with investment, taking the platform into pilot delivery, customer onboarding and wider adoption.
Private beta and investment
Build the next stage
of BRAID with us.
We are speaking with strategic investors and prospective design partners. We plan to work closely with 5 to 10 pilot customers, using their projects to refine the platform and measure its value.
brd.dev ↗