Public Lab
Varun Innovates
Experiments, prototypes, build notes.
Public lab by Varun Siddaraju
XR + AI experiments, notes, and prototypes.
A public lab for applied XR + AI research: experiments, research notes, build logs, and work-in-progress prototypes.
Built by Varun Siddaraju
XR + AI · spatial intelligence · applied systems
50+XR projects and prototypes across education, healthcare, enterprise, architecture, and smart city work.
10+ yearsResearch, product, and delivery experience across XR, spatial computing, and applied AI systems.
2Published books: Apress mixed reality plus Small Teams, Strong Systems.
3+Peer-reviewed publications and research receipts across Sensors, IEEE, and ICDT.
NotesActive research notes across Harmony, XR + AI systems, and multimodal context inference.
Context
Public lab, connected to Varun Siddaraju and VeeRuby.
Varun Siddaraju is the person and research record. Varun Innovates is the public lab. VeeRuby is the company for client delivery.
Current Artifacts
Public artifacts worth inspecting.
Selected frameworks, interfaces, APIs, and notes that show how the research is being translated into practical XR + AI systems.
HarmonyXR_GazePoseContext
A real-time context inference system using gaze, head pose, posture, hands, and spatial context to classify user state as engaged, distracted, transitioning, or idle.
GazeHead PoseHandsBoundaryPosture
OpenSpatialAI API map
An early-stage OpenXR-native platform direction exploring APIs, adaptive behaviors, and world-model experiments for XR workflows.
APIsWorld ModelsOpenXR
XR + AI experiments
A rapid-prototyping lane for WebXR, headset experiments, AI + vision integration, and real-time perception tests.
WebXRAI VisionPrototype Log
Now + Recent
Current work and recently wrapped threads.
Use the tabs to scan active work, published evidence, completed cleanup, and planned public notes without taking in the whole lab at once.
Active Work
The threads currently shaping the lab story and research direction.
Context-aware behavior inference
Modeling gaze, head pose, hands, posture, boundaries, and task context so XR systems can infer human state instead of reacting blindly.
AI command centers and automations
The build lane is active. Current work is turning the patterns, tradeoffs, and examples into clear public notes.
Product modernization pass
Refactoring an existing product while moving it to a newer version, with the goal of cleaner architecture, compatibility, and release-ready behavior.
Published Evidence
Public work and active research notes that support the lab direction.
Small Teams, Strong Systems
The systems book is publicly listed on Amazon and supports the lab's practical systems-thinking thread.
Harmony XR + AI research note
The current note clarifies the systems argument across perception, context, memory, adaptation, and explainability.
Site network redesign
Varun Siddaraju, Varun Innovates, and VeeRuby now separate personal credibility, public lab work, and company delivery more clearly.
Done
Recently wrapped or clarified items that explain why the public lab looks the way it does now.
World models for spatial systems
The current exploration pass is complete and captured as a foundation for continuity, adaptation, and explainability across XR sessions.
SensAI hackathon mentorship
The mentorship is wrapped. The useful residue is sharper thinking around world-model demos, prototype quality, and how teams move from polish to system quality.
OpenSpatialAI and repo cleanup
The repo map now reflects the actual stack: Harmony design work, command centers, books, docs, experiments, and applied XR builds.
Plans
Promising threads that should become cleaner demos, notes, or conversations next.
AI in VR education
Talking with education institutes about integrating AI tutors, adaptive feedback, and richer learning loops into VR education experiences.
Lab Map
One spine: XR + AI systems from research to prototype to product.
Move through the map in order, or jump by intent. Each lane links back to this map and forward to the next lane.
Harmony questions before artifacts.
Research notes hold the argument: how XR systems perceive context, mediate interface complexity, support collaboration, and explain adaptation without hiding control from users.
Harmony framework pass Open
Clarifies perception of context, adaptive mediation, and collaboration orchestration as the three linked layers of the research spine.
Context model boundaries Open
Defines when gaze, head pose, hands, posture, boundaries, and task state are useful enough to drive an adaptive interface.
Evaluation and ethics Open
Keeps adaptation opt-in, reversible, user-tunable, and measurable through task outcomes, cognitive load, trust, and well-being.
Prototypes prove behavior, not polish.
This lane is for small, inspectable demos: adaptive layouts, continuity benches, world-model memory, and AI workflows that reveal what breaks.
Context-aware behavior model Open
Turns multimodal XR signals into interface decisions that can be logged and explained after the session.
Session continuity bench Open
Tests how much spatial memory must persist before an XR session feels continuous instead of reset-heavy.
Failure modes Open
Captures cases where the model reads attention, posture, or intent poorly so the next version can be more trustworthy.
Reusable pieces from the lab.
Tools and SDKs are the reusable surfaces: APIs, plugins, loaders, evaluation helpers, and implementation notes worth using again.
OpenSpatialAI API surface Open
Maps the developer layer for context descriptors, spatial memory, adaptive behaviors, and extension points.
XR interaction utilities Open
Collects reusable patterns for gaze, hands, posture, boundaries, and task-state instrumentation.
Build notes with receipts.
The build log explains what changed, what was removed, what remains blocked, and what needs proof before promotion.
Publish line cleanup Open
Tracks validation, sitemap, build artifact, and deploy-readiness work across the site network.
Artifact promotion rule Open
Strong entries point to screenshots, tradeoffs, and a clear outcome.
From prototype to useful offering.
Product models translate lab patterns into education pilots, command centers, automation systems, and VeeRuby delivery shapes.
AI in VR education Open
Explores AI tutors, adaptive feedback, and learning loops for VR education experiences.
Research-backed services Open
Separates public lab thinking from scoped VeeRuby delivery so both surfaces stay clear.
Lessons from real constraints.
Applied lessons keep production and client patterns abstract enough to be public while still useful for future architecture decisions.
Continuity is not polish Open
If systems forget context between sessions, users repeat orientation work. That is a system design problem, not a UI polish issue.
Latency budgets early Open
XR + AI systems need latency thinking across sensing, inference, orchestration, and rendering before demo polish hides the problem.
Items 1-3
Selected Signals
Archive material is only useful when it sharpens the lab story.
Only the signals that strengthen the current research and prototype story belong here.
3+ peer-reviewed publications plus active research notes
Sensors, IEEE INFOCOM Workshops, and ICDT give the lab real technical lineage instead of pure founder storytelling.
Best Paper Award, ICDT 2018
This is the only award shown here because it has a clean external receipt and ties directly to the research lineage.
Two published books
The Apress mixed reality book and Small Teams, Strong Systems are both public book receipts, with listings on Springer/Amazon.
Research, product, and delivery are connected
Texas State research, Ong product work, and VeeRuby delivery are the trail that explains why Harmony exists at all.
Research + Architecture Notes
Harmony is an early-stage research framework. OpenSpatialAI is an early-stage platform direction.
Harmony focuses on persistent user models, contextual spatial memory, and explainable AI mediation. The lab connects figures, experiments, and implementation evidence to the research direction.
Harmony
XR + AI frameworkSpatial systems that perceive, remember, reason, and adapt across sessions using persistent context rather than single-session triggers.
Notes
Research notes in motionArchitecture arguments, literature notes, figures, evaluation paths, and publication-ready research threads.
OpenSpatialAI
Developer layerAn early-stage GitHub project exploring APIs, memory, adaptive XR behaviors, and world-model experiments built on OpenXR-native primitives.
Evidence
Prototype-backed thinkingResearch ideas connect to working demos, code paths, deployment constraints, and model limits.
GitHub Signal
The repo map is broader than the public profile suggests.
The repo map spans public experiments, research systems, documentation, and product-oriented builds with clear boundaries.
Harmony and OpenSpatialAI stack
Open Repository
The core research and platform layer spans HarmonyXR_GazePoseContext, system design work, and experimental XR + AI prototypes for multimodal context inference and practical adaptive behavior testing.
Command centers, site systems, and operating layers
These repos support AI command centers, team operating systems, and the sites themselves rather than standalone public libraries.
Writing, docs, and personal system repos
Some repos are not product code. They hold books, documentation, personal operating structures, and knowledge architecture that feed the broader lab.
Delivery, demos, and applied XR product repos
This bucket covers company work, immersive products, vertical demos, and partner-facing builds. Useful signal, but not the same thing as reusable open-source infrastructure.
Demo Library
Selected demo evidence, with the full library available.
The selected tab keeps the page compact. Category filters keep the wider public demo trail available.
LabXR Education
AI-assisted VR education environment and one of the clearest bridges between your XR history and current AI direction.
AR Periodic Table Game Education
Interactive chemistry learning in AR.
XR Medical Viewer Healthcare
Medical visualization in immersive space.
Human Anatomy — MR Healthcare
Mixed reality anatomy exploration.
Surgical Instrument Testing Healthcare
Testing surgical workflow concepts in XR.
XR City Model Viewer Architecture
Urban-scale spatial review in XR.
Construction Inspection — MR Architecture
Mixed reality support for site inspection and planning.
AR Anchor: Real Estate Architecture
Property visualization with Unity, ARKit, and Azure Spatial Anchors.
VR Welding Simulation Training
Hands-on welding training for Meta Quest 2.
VR Mining Training Training
Immersive training for mining safety procedures.
VR Driving Simulator Training
Virtual reality driving training on Quest.
Multiuser 3D Model Viewer Enterprise
Collaborative engineering design review in AR.
Metaverse Conference Room Enterprise
Virtual collaboration room built for Quest.
PC Holographic Remote Assist Enterprise
Remote holographic assistance on HoloLens 2.
XR Smart City — Clean Energy Smart City
Spatial computing for renewable energy infrastructure planning.
GIS Data Visualization Smart City
3D globe and geospatial data visualization in mixed reality.
Open Full Library
YouTube channel archive
Commercial Translation
Mature experiments move into VeeRuby delivery.
The lab keeps reusable thinking public. VeeRuby carries scoped projects, enterprise delivery, and client-facing XR + AI outcomes.
Boundary
Public lab here. Client delivery there.
Some ideas from this lab can become client-ready systems through VeeRuby: XR training, mixed reality apps, spatial AI prototypes, Unity/Unreal/WebXR builds, and productized pilots.
Here
Varun InnovatesResearch notes, prototypes, SDK experiments, build logs, product thinking, and applied lessons.
There
VeeRubyEnterprise XR systems, scoped pilots, training simulations, MR apps, technical consulting, and product builds.
Never
No confidential spilloverClient projects can inspire reusable lessons, but names, constraints, and private details stay out.
Signal
Algorithms over anecdotesPublish the reusable method: adaptive UX logic, rendering patterns, AI workflows, evaluation models, and deployment constraints.
Books
Published books that support the lab.
Both books have public listings. The mixed reality book supports the technical XR trail; the systems book supports the operating and builder thread.
Beginning Windows Mixed Reality Programming
Published technical book on hands-on mixed reality implementation.
Open Springer listing
Small Teams, Strong Systems
Published systems thinking book for builders, managers, and founders working under real constraints.
Open Amazon listingContact
For experiments, applied research, collaborations, and serious XR + AI conversations.
Reach out for research collaboration, prototype feedback, talks, or applied XR + AI work. The lab keeps public notes, demos, and references available so the direction is clear before a conversation starts.