ALLSIDES Tech Stack

AI-powered 3D scanning platform converting physical objects into digital twins

Software Development Bressanone, Bolzano 11–50 employees Privately Held

ALLSIDES transforms physical products into photogrammetry-grade 3D digital twins using computer vision and neural rendering. The tech stack reveals a production ML infrastructure: PyTorch, MLflow, Weights & Biases, Kubeflow, and Airflow handle model training and experimentation; differentiable and diffusion-based rendering pipelines are core to their asset generation. Pain points cluster around scaling: quality control at production volume, shipping neural 3D systems, managing multi-terabyte training datasets, and reproducibility across distributed workloads—typical constraints for companies operationalizing research-grade computer vision at e-commerce scale.

Tech Stack 47 technologies

Core StackC++ MLflow Weights & Biases Kubeflow Apache Airflow Prefect PyTorch Docker Delta Lake Linux Python TypeScript React Angular Vue Node.js Django Flask FastAPI GraphQL NVIDIA Substance Painter Blender Horovod Ray DVC NFS Neptune Spring Boot REST+17 more

What ALLSIDES Is Building

◆Challenges

Scaling rendering systems for world-class clients
Scaling quality control at production scale
Shipping neural 3d systems at production scale
Sourcing items that meet scanning criteria
Optimizing procurement costs
Reproducibility across 3d training workloads
Managing multi-tb data storage
Efficient data movement for distributed training
Scaling 3d data platform
Complex workflow integrations

▲Active Projects

Brdf estimation pipelines
Differentiable rendering pipelines
Diffusion-based rendering pipelines
Establish quality standards for 3d scanned assets
Optimize post-production workflows
Document quality standards and best practices
Neural inverse rendering models for geometry and material estimation
Relighting and appearance decomposition pipelines
Generative reconstruction approaches
3d pipeline development

Hiring Activity

Steady15 roles · 6 in 30d

Department

Engineering

Finance

Ops

Seniority

Mid

Senior

Junior

Company intelligence

Find more companies like ALLSIDES by tech stack, pain points and active projects

Get started free

About ALLSIDES

ALLSIDES builds automated 3D scanning systems for brands and e-commerce platforms to generate digital twins of physical products at scale. The company serves apparel, footwear, and outdoor-gear brands requiring mass production of 3D assets for virtual try-on, product photography, and immersive commerce applications. Operations span 3D data capture, neural rendering, and post-production optimization. The engineering-heavy organization (10 of 11 roles) is actively hiring mid and senior-level talent, concentrated in Italy, with focus on rendering pipelines, neural inverse rendering, and quality standardization as production priorities.

HeadquartersBressanone, Bolzano

Company Size11–50 employees

Hiring MarketsItaly

Frequently Asked Questions

What tech stack does ALLSIDES use?

Core ML stack: PyTorch, MLflow, Weights & Biases, Kubeflow, Airflow for training pipelines. Graphics: Blender, Substance Painter. Infrastructure: Docker, Linux, NFS, Delta Lake. Backend: Python, C++, FastAPI, Spring Boot, Node.js. Frontend: React, TypeScript, Vue, Angular.

What is ALLSIDES working on?

Neural rendering pipelines (BRDF estimation, differentiable and diffusion-based rendering), neural inverse rendering for geometry and material estimation, 3D asset quality standards, post-production workflow optimization, and generative reconstruction approaches for digital twin generation.