Senior 3D Technology Engineer

Senior 3D Technology Engineer Digital Twins, Reality Capture & Geospatial Visualization Location: Remote Employment Type: Full-Time Seniority: Mid-Senior candidates welcome About Us We are building a next-generation enterprise digital twin platform at the intersection of reality capture, spatial computing, artificial intelligence and enterprise data. The platform creates intelligent digital representations of real-world assets and environments, connects them with live operational data, and lets users explore, analyze and simulate those environments directly in the browser. The platform already ingests multi-gigabyte reality-capture datasets - LIDAR point clouds, photogrammetry meshes, CAD-derived models - processes them into streamable web formats, and renders georeferenced, site-scale digital twins with live IoT telemetry bound to objects in the scene. Why This Role Exists The platform foundation is built: multi-tenant backend, direct-to-cloud upload of very large datasets, a containerized processing worker, artifact streaming, and WebGL viewers (CesiumJS and Three.js). What determines the quality of the product from here is not more generalist software engineering - it is 3D technology expertise. 3D on the web is a broad discipline: file formats and interchange (glTF, BIM/IFC, CAD, point clouds, meshes), real-time rendering (WebGL, WebGPU, CesiumJS, Three.js), conversion and optimization pipelines, coordinate systems, and streaming performance at scale. We are hiring a senior engineer who is at home in this world - someone with genuine depth in parts of it and the fluency to pick up the rest - to own the spatial side of the platform. The Role You will own the 3D side of the platform end to end: how spatial data of any kind - scans, meshes, BIM models, CAD exports, geospatial layers - is ingested, converted, optimized, and rendered as fast, clean, navigable digital twins in the browser. You will harden what exists, design what is missing, and be the technical authority on everything spatial. You will work alongside platform engineers who own the surrounding infrastructure (APIs, queues, storage, portals), product leadership and domain experts. Where the right answer is a fundamentally different approach, we expect you to say so and prove it with a prototype. What You Will Work On The areas below describe the breadth of the 3D domain. We do not expect any candidate to have worked across all of them - you will grow into several of these areas on the job. Real-Time 3D Rendering on the Web • CesiumJS for georeferenced, site- and city-scale digital twins • Three.js for asset-scale visualization and interaction • WebGL today; WebGPU as it matures • Rendering performance: draw calls, GPU memory budgets, level of detail, streaming large scenes at interactive framerates • Scene composition, camera and navigation UX, picking/selection, in-scene overlays bound to live IoT data 3D File Formats & Interoperability • Web delivery formats: glTF / GLB, 3D Tiles • BIM: IFC and related AEC formats, preserving semantic structure (floors, systems, elements) rather than just geometry • CAD interchange (STEP, DWG/DXF and similar) • Mesh formats: OBJ, PLY, FBX and their quirks (axis conventions, units, materials) • Point clouds: LAS/LAZ, E57 • Compression and optimization: Draco, meshoptimizer, KTX2/Basis textures Conversion & Processing Pipelines • Turning customer-supplied data - in whatever format it arrives - into clean, streamable, correctly placed web content • Mesh processing: decimation, quantization, texture handling, LOD generation • Point-cloud processing: tiling, thinning, classification • Tileset design: LOD hierarchies, geometric error tuning, aggregating per-asset tilesets into seamless site-scale scenes • Automated validation so scale, orientation and placement problems are caught before customers see them Geospatial & Coordinate Systems • Coordinate reference systems (EPSG/UTM), geodetic transforms, local site frames • Vertical datums (geoid vs. ellipsoid heights) for correct globe rendering • Aligning multi-source data about one site - scans, meshes, BIM, drawings - into a single consistent frame • GIS layers, basemaps and geospatial services integration Emerging 3D Technology • Photogrammetry and image-to-geometry (multiple photos → navigable 3D) • Gaussian splatting, NeRFs and other novel capture/rendering approaches as they become production-viable • AI-assisted 3D workflows: dataset classification, drawing interpretation, geometry validation Platform Integration • Processing runs as containerized Python workers on a job queue with S3-compatible storage - you will extend and add capabilities within this architecture, with platform engineers owning the surrounding infrastructure • Defining processing contracts so external engines (e.g. an independent drawing-to-3D service) can plug in cleanly What We Are Looking For Core Requirements • Substantial professional experience building 3D software: real-time rendering, 3D data processing, geospatial visualization, game development, CAD/BIM tooling, or a comparable spatial-computing domain • Solid 3D fundamentals: linear algebra (transforms, rotations, coordinate frames), meshes and materials, scene graphs, level of detail - you can reason about and debug spatial problems from first principles • Working fluency with at least one web 3D stack (Three.js, CesiumJS, Babylon.js or raw WebGL) or deep experience in an adjacent stack (game engines, native graphics) with clear ability to transfer • Strong programming ability in Python and/or TypeScript/JavaScript, with production-software discipline (testing, code review, containers) • The ability to master unfamiliar formats, toolchains and specifications quickly - much of this role is exactly that Good to Have (the more, the better - none individually required) • CesiumJS, 3D Tiles, and georeferenced/globe-scale rendering • BIM/IFC (e.g. IfcOpenShell) or CAD formats and their semantics • Point-cloud tooling: PDAL, Potree, Open3D, laspy or equivalent • Mesh-processing tooling: trimesh, MeshLab, assimp, Blender scripting or equivalent • The glTF toolchain: Draco, meshoptimizer/gltfpack, KTX2/Basis • Coordinate systems and geodesy: PROJ, EPSG registry, datum transforms • GIS stacks: GDAL/OGR, PostGIS, tile servers; ArcGIS ecosystem • WebGPU and shader programming • Photogrammetry (COLMAP, Metashape, RealityCapture), NeRFs, Gaussian splatting • Game-engine experience (Unity, Unreal) or a computer-graphics background • Go (our backend language) • LLM/AI tooling applied to spatial data (classification, validation, extraction) The Mindset We Want • You treat correctness as measurable: a georeference is verified numerically, not eyeballed • You design pipelines that fail loudly, early and with actionable messages - customers upload messy data • You prototype quickly to de-risk ambitious ideas, then engineer the winner properly • You can explain 3D fundamentals to strong engineers who lack graphics backgrounds, and level the team up • You use modern AI-assisted development tools effectively while fully understanding the code produced Example Problems You Would Help Us Solve These are drawn from real problems encountered building the platform: • A customer uploads a folder of site-local OBJ blocks with no CRS metadata, alongside sibling LAS files in UTM. Derive the correct global placement automatically. • A 16-block, 30 GB site scan must stream as one seamless environment. Design the tileset aggregation and LOD hierarchy that keeps it interactive on a mid-range laptop. • Meshes arrive rendered sideways or mirrored. Diagnose and systematize axis-convention handling across formats and converters so it never happens again. • LAS heights are geoid-referenced; the rendering globe is ellipsoidal. Reconcile vertical datums across the pipeline. • Multiple photographs of a physical asset need to become a navigable 3D representation inside a customer's twin. • A pre-flight check must inspect a dataset folder in the browser before upload and predict - reliably - whether processing will succeed, so customers never upload hundreds of gigabytes into a failure. • Simulation engines and IoT telemetry must bind to objects inside the 3D scene and present results visually. Qualifications • 5+ years of professional engineering experience with a substantial focus on 3D technology - graphics, geospatial, BIM/CAD, game development, reality capture or a comparable domain - or equivalent demonstrated depth • A track record of shipped 3D work we can look at - renderers, viewers, pipelines, converters, tools; a portfolio or repository speaks louder than credentials • Experience delivering production software (not only research prototypes or demos) • Excellent written and verbal technical communication • A degree in Computer Science, Geomatics, Mathematics, Engineering or a related field is useful but not required if your experience demonstrates equivalent capability What Success Looks Like Within your first year, we would expect that: • The platform handles a visibly wider range of 3D formats - BIM/IFC and photogrammetry among them - and conversion failures around orientation, scale and placement are engineered out, not patched case by case • Site-scale scenes stream smoothly with proper level of detail, and the in-browser viewing experience feels polished and fast • A pre-flight validation experience prevents doomed uploads before they start • Emerging techniques you championed (better compression, splatting, smarter tiling) have materially improved the product • The engineering team's 3D literacy has visibly risen through your reviews, documentation and mentoring • You are the trusted technical voice on all things spatial within the company