Straight answers about Laan Labs: who we are, what we build, how technology licensing works, and the 3D capture and spatial computing technology behind our products. If your question is not here, email labs@laan.com.
Last updated October 10, 2026
Laan Labs is a software development studio that specializes in computer vision, spatial computing and AI. Founded by brothers Christopher and Jason Laan, the company has been building next-generation technology for almost 20 years, both for clients and as its own products. Laan Labs apps have been downloaded more than 70 million times, and the company has worked with brands including Nike, HBO, Toyota and MLS.
Laan Labs was founded by brothers Christopher Laan and Jason Laan. Christopher is based in New Orleans and Jason in New York. Both come from engineering backgrounds in electrical and chemical engineering and apply that rigorous, scientific approach to software. Together they have led the teams behind well-known products such as Face Swap Live, iVideoCamera, 3D Scanner App and early work on Venmo. Projects at Laan Labs are founder-led, so clients work directly with them rather than through layers of account management.
Laan Labs is based in New York and New Orleans, Louisiana, and works remotely with clients across the United States and internationally.
Laan Labs is best known for groundbreaking mobile app development. Over the years the company has had a large number of number one apps in the App Store, and its apps have been downloaded more than 70 million times. Landmark apps include iVideoCamera, which brought video recording to early iPhone models that shipped without it; Face Swap Live, the original real-time face swapping app, released in 2015; and 3D Scanner App, one of the first and most popular apps to use the LiDAR sensor on iPhone and iPad. More recently the company has focused on Gaussian splatting, mobile 3D capture and spatial computing, with products such as 3D Splat App for Mac, RadianceView for inspection and documentation, and AirDraw for Apple Vision Pro.
Laan Labs has worked with both large brands and small companies, including Nike, HBO, NBC, MTV, TBS, Toyota, Volkswagen, Prada, Vimeo, Twitter, Grubhub, Venmo and MLS, as well as startups, insurers and industrial clients that license its 3D capture technology.
Laan Labs technology has been used in insurance, medical, construction, manufacturing, retail, aerospace, automotive, real estate, media and entertainment. Typical projects include 3D capture and measurement of spaces and objects, inspection and documentation of physical assets, computer vision for photography and e-commerce, and AI features inside consumer mobile apps. See the industries page for insurance, real estate, medical and body scanning, robotics, utilities, aerospace, construction, automotive, manufacturing and industrial inspection, media, VFX and games and art, museums and cultural heritage in detail.
Laan Labs offers four main services: AI solution development, from consulting and proof-of-concepts through production-ready systems; computer vision and spatial computing, including custom CV pipelines, 3D capture and XR applications; rapid prototyping that validates an idea with a working prototype in weeks; and technology integration that connects AI, machine learning and computer vision to a client's existing products and systems. Each is described on the services page.
Laan Labs builds custom computer vision systems for object detection and tracking, pose estimation, segmentation, measurement, 3D reconstruction and scene understanding, with a particular focus on running models on-device on phones, tablets and headsets rather than only in the cloud. Examples include real-time 3D car pose estimation for automotive photography, person segmentation for masking 3D scans, and extracting floor plans and semantic geometry from room scans.
Yes. Laan Labs integrates and fine-tunes large language models and combines them with computer vision, for example in CreatorView, a system that watches social media video and uses computer vision plus LLMs to assess creator and brand fit. Laan Labs also builds AI agents and LLM-powered features into existing products.
Yes. Laan Labs develops for Apple Vision Pro and visionOS. It published AirDraw, a 3D finger-painting app for Vision Pro, and uses the headset in client work such as aerospace parts inspection and documentation, where spatial computing lets an inspector review 3D captures and photographs in place. See the Apple Vision Pro blog post and the aerospace inspection case study.
Yes. Laan Labs generates synthetic training data when real annotated data is expensive or unavailable, rendering varied scenes with 3D assets and automatic labels. It has used this approach to train car pose estimation and foot pose estimation models at a fraction of the cost of manual annotation. See the car dataset and foot dataset case studies.
A typical Laan Labs engagement starts with a short scoping conversation, followed by a working prototype or proof-of-concept in a matter of weeks. Production systems follow from there in weeks or months rather than years, depending on scope. Laan Labs is a small, senior team, so there is little overhead between a decision and working software.
Projects at Laan Labs are founder-led. Christopher and Jason Laan are directly involved in the engineering, supported by a small team of experienced engineers. Clients have direct access to the people building their software, with no layers of project management in between.
Yes. Laan Labs works with both. Large clients have included Nike, HBO, Toyota and Volkswagen; startup clients have licensed source code to accelerate their own computer vision and machine learning teams, or engaged Laan Labs to build a first prototype and validate an idea.
Laan Labs builds native apps for iOS, iPadOS, macOS and visionOS, as well as Android, web and cloud back ends. Its specialties include Apple frameworks such as ARKit, RoomPlan, Metal and Core ML, GPU programming in Metal and CUDA, 3D reconstruction tools such as COLMAP and Gaussian splatting, and machine learning in PyTorch. The company has full-stack capability from mobile apps to cloud infrastructure.
Yes. Laan Labs licenses its apps and technologies as white-label and source code solutions. Its 3D capture pipelines for LiDAR, TrueDepth, RoomPlan and photo-based reconstruction are available both as frameworks that your team integrates and as source code that your team can build on, which greatly reduces product development cost and time to market.
A framework license from Laan Labs provides a compiled SDK that your developers integrate into an app, with Laan Labs maintaining the technology inside it. A source code license provides the full source so your own engineering team can modify and extend it. Source licenses suit companies with in-house computer vision or machine learning teams who want a proven starting point; see the source code license case study.
Yes. Laan Labs delivers end-to-end solutions built on its technology as well as plain licenses. For example, it built a custom licensed version of 3D Scanner App for an insurance company to use in policy estimation and claims, improving both efficiency and accuracy. See the insurance workflow case study.
Licensing cost depends on the technology, the scope of use and whether it is a framework, source code or custom integration. Contact Laan Labs at labs@laan.com with a description of your use case for pricing and implementation options.
Gaussian splatting is a technique for turning ordinary photos or video of a real scene into a photorealistic 3D scene that renders in real time. Instead of a polygon mesh, the scene is represented as millions of small, soft, overlapping 3D blobs (Gaussians), each with a position, shape, color and opacity, which are optimized until renders match the original photographs. Radiance fields are a Laan Labs specialty: it builds Gaussian splatting tools such as 3D Splat App, uses splats in RadianceView, and offers Gaussian splatting services. There is a plain-language primer at What Is Gaussian Splatting?
All three reconstruct 3D from photographs. Photogrammetry produces a textured polygon mesh, which is editable and measurable but struggles with reflective, thin or finely detailed surfaces. Neural radiance fields (NeRFs) render very realistic novel views, but early versions were slow to train and even the fast variants are demanding to display. Gaussian splatting matches NeRF image quality while rendering in real time, which makes it practical for interactive viewers, inspection and web delivery. Laan Labs compared these approaches on real assets in its InfraTech 3D scanning case study.
Mobile LiDAR scanning uses the LiDAR sensor built into iPhone Pro and iPad Pro models to capture rooms, buildings and large objects as 3D meshes and point clouds in minutes, with no extra hardware. It is well suited to room-scale measurement, documentation, floor plans and visualization, with accuracy typically in the centimeter range under good conditions; it is not a replacement for survey-grade instruments where millimeter precision is required. Laan Labs pioneered this workflow with 3D Scanner App and continues it in OpenSurvey3D and RadianceView. See how mobile LiDAR is redefining professional 3D workflows.
Yes. Laan Labs built a volume measurement tool into its mobile LiDAR scanning pipeline: the operator walks around a pile of gravel, dirt, sand, mulch or other bulk material with an iPhone Pro or iPad Pro, outlines the base of the pile on the scan, and reads the volume in cubic meters, cubic yards and cubic feet. The same tool measures a hole or trench to give the fill required. Tonnage follows from a density factor per material. Phone LiDAR suits piles up to a few truckloads; larger piles and whole yards are captured with photogrammetry or a drone and measured the same way. See the stockpile volume measurement case study.
Yes. Laan Labs built a job estimating app for outdoor work on its own scanning pipeline: an operator scans the work area with an iPhone, and the app measures the quantities that drive the job, such as the volume of a stump above grade, the area and depth of a bed, cut and fill for grading, and lengths of trench or edging. Guided questions capture what a scan cannot see, such as species, condition, access, slope and utilities, and an estimating model turns quantities and conditions into hours, materials to order and a customer-ready quote. The app is available to partners under license rather than in a public app store. See the job estimating app case study.
TrueDepth scanning uses the infrared depth sensor on the front of the iPhone, the same sensor used for Face ID, to capture small objects and body parts in fine detail at close range. Because it works entirely on the device, it suits applications where photos cannot be sent to a cloud reconstruction service, such as medical and custom-fit products. Laan Labs has licensed TrueDepth scanning pipelines to organizations that need precise small-scale scans; see the TrueDepth scanning case study.
RoomPlan is an Apple framework that uses the iPhone and iPad LiDAR sensor to produce a parametric model of a room: walls, doors, windows and furniture as labeled geometry rather than a raw mesh. Laan Labs builds on RoomPlan for floor plan extraction and semantic scene understanding, and its open-source OpenPlan3D floor plan editor imports RoomPlan scans directly. See the blog post on semantic geometry with RoomPlan.
RadianceView is an inspection-grade visualization platform built by Laan Labs. It brings mobile 3D scans, traditional LiDAR, Gaussian splats and high-resolution photography together into a single visual record that opens in a web browser, so teams can inspect, measure and report on an asset without being on site, with every finding tied to the place it was observed. It has been used to document end-of-life airliners for parts reuse; see the aircraft documentation case study.
Laan Labs has published apps with more than 70 million combined downloads, including 3D Scanner App, Face Swap Live, AirDraw for Apple Vision Pro, 3D Splat App for Mac, Air Measure, Tap DJ, Tap Roulette, and earlier apps such as Sonar Ruler, iVideoCamera and iTimelapse. Current free and open-source projects include OpenPlan3D, OpenShape3D, OpenWater, OpenSurvey3D and SpaceDust. See the full list on the apps page.
3D Splat App is a free Mac app from Laan Labs that trains Gaussian splats locally on Apple Silicon. It imports photos, COLMAP datasets or Nerfstudio projects, recovers camera poses with a built-in COLMAP, trains on the Mac GPU using Metal with no cloud upload, and exports PLY, compressed PLY, SPZ and SOG. It also works as a splat viewer and editor, with crop, paint-select and transform tools, and opens splats trained elsewhere. It requires an M1 or later and macOS 15.6 or later.
Laan Labs created 3D Scanner App, one of the first LiDAR scanning apps on iOS, and no longer operates its App Store listing, which was sold to a private investor in July 2025. The App Store listing for Air Measure was sold in 2024. In both cases only the listing was sold: Laan Labs retains the intellectual property rights to the underlying technology, continues to build 3D capture products including RadianceView, OpenSurvey3D and 3D Splat App, and licenses its 3D capture pipelines to other companies. For custom solutions built on this technology, contact us.
Yes. Laan Labs maintains several open-source projects on GitHub: OpenPlan3D, a floor plan editor that imports iPhone RoomPlan scans; OpenShape3D, a direct-modeling 3D CAD app for iPad built on the OpenCASCADE kernel; OpenSurvey3D, which combines 3D scanning, GNSS surveying and AR asset finding for field work; and OpenWater, a GPS speed tracker for wingfoiling, windsurfing and sailing. See other apps for the full list.
AirDraw is a Laan Labs app for Apple Vision Pro that lets you paint in 3D using your fingers, with your room as the canvas. It includes Sculpt, Physics and Voxel drawing modes and can export creations as USDZ files.
Email Laan Labs at labs@laan.com or use the contact form with a short description of the project, the platform and the timeline. You will hear back directly from one of the founders.
Yes. Laan Labs regularly signs non-disclosure agreements before discussing confidential projects, and much of its client work is under NDA.
Yes. Laan Labs works remotely with clients across the United States and internationally, with on-site visits when a project calls for it, such as capturing a facility or asset in person.
Yes. Technology integration is one of Laan Labs' core services. The team can audit an existing computer vision or machine learning system, improve its accuracy or performance, move models on-device, or connect AI capabilities to existing apps and back ends.
Email labs@laan.com or use the contact form and one of the founders will get back to you.