Nebula Cloud

09/02/2026

From raw imagery to operational intelligence — in one governed pipeline.

Nebula Cloud Studio just powered a real, end-to-end GeoAI prototype: raw drone imagery in, live operational WebGIS out.

🛰️ AI detection & classification on high-res orthoimagery
📐 Raster → GIS-ready vector features
🗺️ Live WebGIS with an AI analyst for natural-language queries
🔍 Full traceability — confidence scores, provenance, human validation

3,000+ georeferenced features extracted end-to-end. A working system, not a mockup.

Agents flag and hand off, humans decide.

The same GIS + remote sensing + digital twin engines also power heritage digital twins, civic mapping, municipal intelligence, and public geo-data cataloguing.

One governed core. Many domains.

Explore → nebulacloud.studio

09/01/2026

We’ve published an open interactive showcase on the 2026 Nepal flash floods — a 3D Digital Twin & What-If Scenario Lab for the Bhote Koshi–Trishuli corridor.

It lets you replay a representative event, explore the terrain in 3D, and test what-if scenarios by changing breach conditions, rainfall, debris, river flow and infrastructure assumptions.

Live demo:
https://lnkd.in/dR_yyJiN

Open-source repo:
https://lnkd.in/dT2eRe4z

This is a scenario-based research simulation, not an official flood warning or evacuation tool.

Would really appreciate feedback from anyone working in GIS, hydrology, disaster management, remote sensing, humanitarian mapping or digital twins.

Nepal Live ReportersNepal Nepal Telecom Nepal Police

08/29/2026

We’re opening Nebula Cloud Labs to academia and research institutions. 🔬

A platform for researchers, universities and R&D teams to access AI, GPU computing, geospatial, engineering, simulation and research-ready workbenches while connecting with the applied research we are building at Nebula Cloud.

Research. Experiment. Validate. Demonstrate.

If you’re a faculty member, researcher, student team or institution looking to accelerate research, I’d love to connect.

👉 Explore Nebula Cloud Labs:
https://lnkd.in/dEqiHxW9

08/25/2026

Where should someone get building footprints for Telangana?

A government source? OpenStreetMap? Microsoft’s open building footprints? Google Open Buildings?

Each can be useful but they are not interchangeable.

The more important questions are:

Which source actually covers the geography? Is it accessible? Is the geometry usable? How current is it? What license applies? Is there an authoritative alternative? And what happens when none of the available sources is good enough?

We kept running into variations of this problem across geospatial work.

So we built Nebula Civic Index — GeoIndia.

GeoIndia is an open geospatial data intelligence initiative from Nebula Cloud Public Good Labs, built around a simple workflow:

Find → Verify → Compare → Act

It brings public datasets, services and portals together with intelligence around provenance, access, GIS readiness, source alternatives and evidence gaps.

The underlying datasets remain with their original publishers. GeoIndia is an independent discovery and intelligence layer not an official government portal and not a data host.

When discovery needs to move into data preparation, analysis, transformation or ex*****on, Nebula Cloud Studio provides the next layer.

We’d particularly like to connect with researchers, think tanks, geospatial organizations and institutions working with similar public-data challenges.

Explore GeoIndia:
https://geoindia.nebulacloud.in

Find the data. Know whether to trust it. See what’s missing.



Built with Nebula Cloud Studio

08/20/2026

From a static 3D model file → a fully interactive browser dashboard. No CAD software, no install, no plugins.

We took a production-quality concept car asset (glTF/GLB, PBR materials, clearcoat paint) and turned it into a real-time interactive viewer — built and deployed through Nebula Cloud Studio.

What it does:

🎨 Switch between 3 factory paint finishes in real time

🚪 Open/close scissor doors, hood, and rear hatch with smooth animation 💥 Explode view to inspect body panels and construction

🔄 Turntable mode, free orbit, and 5 camera presets (front, back, side, top, interior)

📸 Live wireframe toggle, lighting control, and one-click screenshot export

All of it runs in a browser — zero backend, zero build step, fully responsive on mobile.

This is the kind of deliverable design teams, product managers, and marketing teams need for stakeholder reviews and client demos, without depending on specialist CAD software or platform-specific viewers. Point someone at a URL, and they can inspect material variants, mechanical articulation, and detail — no software required.

🖥️ Live demo → https://lnkd.in/gs3V4P-b

📦 Repo → https://lnkd.in/gPg7jy3C

⭐ If this is useful, a star on the repo helps more people find it.

Want a similar interactive deliverable for your own 3D models, CAD assets, or product configurator? Let's talk → https://nebulacloud.studio

08/19/2026

.studio — Municipal Property Intelligence is a map-first Web GIS and AI-assisted evidence layer designed to help urban local bodies reconcile physical building reality with municipal property records.

For our Tirunelveli demonstration, Studio combines independent building and temporal evidence from multiple sources and converts them into inspectable, provenance-aware geospatial intelligence.

The current public demonstrator contains:

🏢 36,069 building-evidence clusters generated through multi-source conflation

🔎 14,471 structures independently corroborated by Google + Microsoft building datasets but absent from OSM

⚠️ 3,940 cases requiring geometry/evidence review rather than forcing uncertain matches

🛰️ 2016–2023 temporal building intelligence for understanding persistence and built-environment change

🗺️ Settlement-intensity and municipal-reconciliation context generated from openly licensed building/temporal evidence

But the most important part is what the system does not do.

Studio does not autonomously declare a property unassessed, calculate speculative tax, or replace municipal authority.

The workflow deliberately stops at the correct institutional boundary:

Physical + temporal evidence
→ reconciliation candidate
→ GIS / field verification
→ municipal officer decision
→ existing statutory assessment workflow

Once authorized municipal data such as PTIN/property registers, building permissions, current ward/zone GIS and DCB records are connected, the same evidence framework can support controlled property-record reconciliation and verification.

See the work yourself

🌐 Live Web GIS:
https://studio-public-demos.github.io/municipal-property-intelligence/demo/

💻 Public project showcase:
https://github.com/studio-public-demos/municipal-property-intelligence

08/18/2026

A BIM model isn't just geometry.

It contains walls, slabs and beams — but also GlobalIds, materials, property sets, quantities and spatial relationships.

The challenge is that much of this information lives inside heavyweight BIM formats, while downstream web, mobile, XR and digital-twin applications need lightweight geometry and structured, queryable data.

So we used NebulaCloud Studio to build a small open-source proof of concept:

IFC / Revit-derived BIM → GLB/GLTF + structured metadata.json + REST API

The workflow:

→ ingest BIM data
→ parse geometry + engineering semantics
→ generate optimized GLB/GLTF
→ preserve structured BIM metadata
→ explore the result interactively in 3D
→ compare two discipline models
→ expose the outputs through APIs

One of the more interesting parts is the model comparison workflow.

Using the Architecture and Structural views of the same building, the POC identifies:

4 common · 14 added · 15 removed elements

And on the architecture sample:

226 KB IFC → 17 KB GLB — approximately 92% smaller

while the structured metadata preserves information required for downstream applications.

The simplest way I think about it is:

The GLB tells an application what the building looks like.
The metadata tells it what the building means.

That combination can become a foundation for applications such as:

• web and mobile BIM viewers
• digital twins
• design coordination and change intelligence
• construction-tech platforms
• asset/facility applications
• AR/VR/XR experiences
• engineering data APIs

This is intentionally a proof of concept and reference implementation, not a production BIM platform.

And rather than keeping the code behind a demo, we're making the reference implementation available under the MIT license.

Try it, fork it, modify it or build on it →
https://github.com/studio-public-demos/bim-cloud-pipeline

The bundled IFC workflow runs locally with real buildingSMART sample data and does not require a BIM desktop license or cloud account.

And if your organization has a BIM, digital-twin, engineering or domain-specific workflow that could benefit from a similar approach, I'd be interested in discussing the problem.

Built with NebulaCloud Studio — an example of going from a domain engineering requirement to a working application.

08/17/2026

We’ve added another capability layer to the Nebula Cloud Workbench — an integrated 3D Reconstruction, NeRF and Gaussian Splatting stack for demanding reality capture and spatial AI workloads.

Current 3D/GS stack
🔹 COLMAP — Structure-from-Motion and camera pose estimation
🔹 Nerfstudio — NeRF & Gaussian Splat training with live visualization
🔹 SuperSplat — Edit, optimize and publish trained splats
🔹 Meshroom — GUI-based photogrammetry and reconstruction

End-to-end workflow
Photos → COLMAP → Nerfstudio → Export PLY/Splat → SuperSplat → Interactive 3D experience

The objective is bigger than installing individual tools.

We are building a unified workbench where 3D reconstruction, geospatial intelligence, computer vision, engineering simulation and AI agents can work together — with deeper orchestration through Nebula Cloud Studio

Because these workloads vary significantly, we also provide a choice of CPU, GPU and storage configurations based on dataset size, computational complexity and performance requirements. NVIDIA GPU's from T4 - H100 on PAYG basis.

This allows the infrastructure to scale from smaller experiments to demanding workloads involving:
• NeRF & Gaussian Splatting
• Photogrammetry and drone imagery
• Large geospatial datasets
• Digital twins and BIM
• Industrial 3D reconstruction
• Robotics & Physical AI environments
• Engineering and simulation workflows

Who can use this?

🏗️ AEC, infrastructure & utilities — site capture, asset documentation and digital twins

🛰️ Geospatial, drone & surveying teams — reality capture and 3D spatial intelligence

🏭 Manufacturing & engineering companies — plants, equipment and industrial assets

🏛️ Museums, archaeology & heritage organisations — digital preservation

🤖 Robotics, Physical AI & simulation teams — realistic 3D environments

🎓 Universities & R&D labs — applied research and experimentation

☁️ Cloud, GPU, storage, satellite, drone & 3D technology providers — integration and co-development

🤝 System integrators & consulting firms — joint customer solutions

Have photos, drone imagery, a site, facility, physical asset or large 3D dataset you want to test?

👉 Talk to us about a pilot, POC, research collaboration, technology partnership or commercial deployment. We can configure the compute and storage footprint around your workload — rather than forcing the workload into a fixed environment.

08/15/2026

08/15/2026

From raw Census data to 5 publication-ready thematic maps.

The entire pipeline ran end-to-end through Nebula Cloud Studio fully automated, zero manual GIS work.

We took district-level Census of India 2011 data — 642 districts — and turned it into a complete geospatial deliverable: five choropleth maps covering Literacy Rate, S*x Ratio, Scheduled Caste & Tribe population, and Workforce Participation.

Every map is classified with 6 quantile classes, laid out at A4 / 200 dpi with legends, scale bars and attribution — then packaged as high-res PNGs, individual + combined PDF reports, an interactive web gallery, and a demo video.

This is what Nebula Cloud Studio does: GIS data visualization that turns raw tables into decision-ready, shareable outputs for foundations, researchers and policy teams.

🗺️ Live demo → https://lnkd.in/dSFwQTAU
📦 Repo → https://lnkd.in/d2XQb5vB

⭐ If this is useful, a star on the repo helps more people find it.

Want a similar deliverable for your own data? Let's talk → https://nebulacloud.studio