AI, 3D & Computer Vision Portfolio

From agentic AI to 3D pipelines — custom software that removes real bottlenecks.

Rendering backends, computer vision pipelines, enterprise modernization, VR applications, and AI-driven automation — built for technology platforms, manufacturers, and industrial brands.

View our work ↓
Our Work

Selected Projects

Ten projects across rendering, computer vision, enterprise AI, VR, and industrial automation. Each one starts with the workflow, not the technology.

AI-Driven Legacy Code Conversion
Project 01
AI-Driven Legacy Code Conversion
Agentic AI workflows that convert hundreds of legacy screens to modern architecture. Goal: client self-sufficiency.
3D Product Configurator
Project 02
3D Product Configurator
Three.js assembly engine for 20+ component products. Real-time compatibility validation, conflict detection, and AR export.
Photorealistic Rendering Backend
Project 03
Photorealistic Rendering Backend
Ray-traced rendering pipeline integrated into a visual content platform. Photorealistic stills, 360-degree spinners, and video — with fast preview and full-quality modes.
AI + 3D-Printed Automotive Parts
Project 04
AI + 3D-Printed Automotive Parts
Image-to-3D-printable-part pipeline for a global car manufacturer. Manufacturability constraints enforced automatically.
Synthetic Dataset Generation + CAD Detection
Project 05
Synthetic Data + CAD Detection
No real training data? We built the entire pipeline — procedural 3D models, synthetic rendering, and YOLO training. 90%+ mAP.
Prompt-Based CAD Editor
Project 06
Prompt-Based CAD Editor
Text commands that edit STEP CAD files programmatically. Fully on-premise LLM — no cloud dependency, no data leaves the building.
VR Room Planner
Project 07
VR Room Planner
Meta Quest 3 application for configuring furnished rooms in VR. Add, move, and swap furniture with live pricing and automatic reports.
Medical Document Verification
Project 08
Medical Document Verification
Hebrew OCR pipeline for medical referral forms. Multi-engine benchmarking. On-premise for data privacy. Under 13 seconds per document pair.
Virtual View Camera Calculation
Project 09
Virtual View (6DOF Spatial Parallax)
Physically accurate virtual camera from real-time head tracking. SDK extension that turns a display into a window into a virtual world.
Procedural 3D Model Generation
Project 10
Procedural 3D Model Generation
10,000 realistic flatware 3D models from a procedural generator. Infinite unique variations, production-ready output, fully automated.
About

3D Foundry Labs

3D Foundry Labs is a bespoke R&D software firm. We build custom automation tools by combining AI, computer vision, 3D graphics, and software engineering into systems that solve real operational problems.

Every engagement starts by understanding where time and money are being lost — then we build the software to remove that bottleneck. We do not offer off-the-shelf products. Every tool we build is designed to behave the way your team already works.

Eliran Dahan
Eliran Dahan
CEO & Co-founder

20+ years developing large-scale software and imaging systems used by tens of millions of users. B.Sc. Electrical Engineering, Technion — Israel Institute of Technology. Co-inventor on 4 US patents.

LinkedIn
Tal Kenig
Tal Kenig
CTO & Co-founder

20+ years in computer vision, machine learning, and imaging systems. Built advanced vision platforms at GE Healthcare and Samsung Electronics. M.Sc. Electrical Engineering (Summa Cum Laude), Technion. Co-inventor on 9 US patents.

LinkedIn