Skip to content
AGL — Aidan G. Larson logo

3D Artist · Unreal Developer · VR/AR

Aidan Larson

3D artist in Philadelphia building real-time worlds in Unreal Engine 5. I model, texture, light, and ship VR experiences, and I test AI tools inside the pipeline.

View workShading viewer
Scroll

Selected work

Graduating June 2027 · Open to VR, AR, and real-time 3D rolesResume PDF ↗

01 — About

Topology first, atmosphere always.

I'm a 3D Visualization & Digital Designer and VR/AR specialist based in Philadelphia, finishing a B.S. in Digital Media and Virtual Production at Drexel University (Dean's List).

My pipeline runs full-stack: modeling and sculpting in 3ds Max, Maya, and ZBrush, texturing in Substance Painter and Designer, and real-time build-out in Unreal Engine 5 — shipped to headset at a locked frame budget.

On occasion, I build AI directly into that pipeline: custom Blender and Maya MCPs, physics-based animation in Cascadeur, and Claude Code with MCP — including a custom Unreal MCP integration — for in-engine scripting and agentic modeling workflows.

Toolset

3D & Sculpt
MayaMaya3ds Max3ds MaxZBrushZBrushBlenderBlenderAutoCADAutoCADRhinoRhinoRevitRevit
Texturing
Substance PainterSubstance PainterSubstance DesignerSubstance Designer
Real-time
Unreal Engine 5Unreal Engine 5UnityUnityUEFNUEFN
Compositing / VFX
NukeNukeHoudiniHoudini
Capture
Motion captureMotion capturePhotogrammetryPhotogrammetry
Agentic tooling
Claude CodeClaude CodeChatGPT CodexChatGPT Codexcustom Unreal MCPcustom Unreal MCPMaya MCPMaya MCPBlender MCPBlender MCPHunyuan ProHunyuan ProCascadeurCascadeurNano BananaNano BananaFluxFluxMidjourneyMidjourneyNVIDIA DLSSNVIDIA DLSS
Nissan S13 wireframe render passNissan S13 normals render passNissan S13 textured render pass

02 — Study · Facility / VR

Cobbs Creek Golf Facility

Architectural visualization and a standalone Quest 3 putting simulator built at EwingCole — 3ds Max to Unreal Engine 5 via Datasmith, packaged at a locked 60fps for client-facing design review.

Skills used

  • Real-time VR development
  • Blueprinting
  • Modeling
  • Texturing
  • Environment design
Brief
Architectural visualization of the facility, plus a putting simulator clients could try in headset during design review.
Constraint
A standalone Quest 3 build, so the scene had to hold a locked 60fps on mobile hardware.
Decision
3ds Max to Unreal Engine 5 through Datasmith, with Blueprints driving the simulator.
Result
Owned the full AR/VR pipeline as my EwingCole co-op capstone deliverable, then packaged the build for Meta Quest 3.

03 — Study · Horology

Citizen Eco-Drive Study

A full product visualization pass on a Citizen classic — skeleton movement, guilloché dial, calfskin strap and engraved caseback, broken down into its render passes from wireframe to final PBR. Hover or drag across the frame to drop a loupe onto the surface and inspect the micro-detail.

Skills used

  • Modeling
  • Texturing
  • Lighting
  • Rendering

frame data

lens
50mm · f/4
shader
brushed steel · sapphire
plate
tiled platform

plates

render passes

04 — Capstone · VR Experience

Through the Fog

My senior capstone: a VR recreation of the initial frontline assault on the Cliveden mansion in Germantown, Philadelphia — October 4, 1777. Step through the phases of the engagement as the player would move through it in headset.

Skills used

  • Real-time VR development
  • Modeling
  • Texturing
  • Lighting
  • Environment design
  • Animation
  • Blueprinting
  • Agentic systems testing
Brief
A 30 to 60 second first-person VR combat scene recreating the Continental assault on Cliveden, October 4, 1777, for Meta Quest 3.
Constraint
A 500,000 triangle budget for the whole scene, targeting Meta Quest 3.
Decision
Model most assets by hand from a site visit, the 1776 floor plan, and period paintings treated as secondary evidence. Claude works through Blender for many of the model updates, and every AI tool used gets logged along with where it fell short.
Status
In production. Scene, 30-soldier formation, and firing animations are running in Unreal; the VR build phase is underway.

Agentic build

Agents in the pipeline

For this project I experimented with building custom MCPs for Maya, Blender, and Unreal Engine — driven from Claude Code and GPT Codex. That unlocked automated systems that assisted in building parts of the immersive experience, from modeling, rigging and retopology through to in-engine build, scripting and blueprinting.

  • Maya MCP

    modeling, rigging & retopology

  • Blender MCP

    modeling, rigging & retopology

  • Unreal MCP

    in-engine build, scripting & blueprinting

Through the Fog — Cliveden assault VR scene, Artillery TrainThrough the Fog — Cliveden assault VR scene, The ColumnThrough the Fog — Cliveden assault VR scene, Horse TeamsThrough the Fog — Cliveden assault VR scene, Under the OakThrough the Fog — Cliveden assault VR scene, Bivouac

A-01

Artillery Train

05:00 · Germantown Road

Horse teams hauling light guns and a supply wagon down the Germantown road in the pre-dawn fog.

Current VR Testing

1777 Brown Bess Musket Reload in Unreal Engine 5.8

0:000:00

First-person reload test in Unreal Engine 5.8: bare hands reach for the Brown Bess and work the lock step by step, with VR interaction markers anchored to the moving parts of the musket.

05 — Study · Environment Art

Abandoned Nissan S13

A Nissan 180SX left to the ivy inside a collapsing corrugated shed. The set was built and dressed once, then lit twice — a hard practical work-light rig at night and a soft overcast pass by day. Switch the lighting state and step the shot list.

Skills used

  • Modeling
  • Lighting
  • Texturing
  • Foliage
  • Environment design
Brief
Final project for a Texturing and Lighting course: an abandoned 180SX in an overgrown corrugated shed, kept on as a portfolio piece.
Decision
Build and dress the set once, then light it twice, a hard work-light rig at night and a soft overcast pass by day.

shot 01 / 03

S13 diorama environment render — Work Light Three-QuarterS13 diorama environment render — Door GapS13 diorama environment render — Rear RevealS13 diorama environment render — Overgrowth CloseS13 diorama environment render — Head-OnS13 diorama environment render — Establishing Wide

N-01

Work Light Three-Quarter

35mm · practical rig · 1/60

Two halogen stands do all the work — hard speculars on wet paint, everything past the shed falling to black.

06 — Study · Consumer Audio

JBL Flip Study

A materials-first product build: woven mesh, soft-touch rubber and machined trim. Scroll through the section and the control deck wipes across the body render along a diagonal seam — or lock a plate below to read the shading notes.

Skills used

  • Modeling
  • Texturing
  • Lighting
  • Rendering
Brief
A product study for a course: model and texture a JBL Flip from reference.
Constraint
One week, start to finish.
Decision
Spend the time on materials: woven mesh, soft-touch rubber, and machined trim, modeled in Maya and textured in Substance Painter.
Scroll-driven wipe
JBL Flip product visualization — body and grilleJBL Flip product visualization — control deck

JBL-01 · Body & Grille

07 — Study · Character Facial Design

Human Realism Design

A likeness study taken from sculpt to shaded turnaround — pore-level skin detail, a strand groom rebuilt as cards for headset, and refractive eyes that hold up in profile. Drag across the plate to rotate the head, or step to a locked view.

Skills used

  • Modeling
  • Sculpting
  • Hair simulations
  • Texturing
  • Rendering
Digital human character study — front viewDigital human character study — three-quarter viewDigital human character study — profile view

T-00

YAW 00°

drag to turn

08 — Contact

Let's build something in real time.

Graduating June 2027 · Open to VR, AR, and real-time 3D rolesResume PDF ↗

agl55@drexel.edu

Elsewhere

Opens your email app with the message ready to send.

MUSTANG · FINAL LIT PASS

Ford Mustang — final lit render with floor reflection
AGL logo

© 2026 Aidan Larson — Philadelphia, PA

Back to top ↑