Nexxa.ai Tech Stack

AI agents for industrial engineering workflows in heavy industries

Software Development Sunnyvale, CA 11–50 employees Founded 2024 Privately Held

Nexxa builds specialized AI agents targeting industrial engineers in railroads, energy, construction, and mining—sectors where fragmented legacy tech stacks slow down critical repetitive work. The stack is deep in ML infrastructure (PyTorch, TensorFlow, TensorRT, ONNX, DeepSpeed) and computer vision (CLIP, Vision Transformers, OpenCV), with all 9 active roles filled by senior engineers, suggesting a lean, technical-first team focused on shipping production systems rather than sales motion.

Tech Stack 27 technologies

Core StackPyTorch TensorFlow Python C++ AWS Docker Kubernetes Jira GitHub Go Hugging Face TensorRT ONNX OpenVINO TorchScript GCP Azure CLIP CI/CD JavaScript/TypeScript OpenCV JAX Ray Lightning DeepSpeed Vision Transformers GANs

What Nexxa.ai Is Building

◆Challenges

Adopting generative ai in legacy environments
Legacy system dependencies
Rising operational costs
Workforce transitions

▲Active Projects

Real-time inference systems
Object detection and tracking
Ai and computer vision program
Multimodal llm/vlm integration
Genai and cv solutions for customer problems
Large-scale ml training workflows
Custom ai solutions for manufacturing customers
Ai programs
Software engineering programs
Custom genai solutions

Hiring Activity

Minimal9 roles · 0 in 30d

Department

Engineering

Seniority

Senior

Company intelligence

Find more companies like Nexxa.ai by tech stack, pain points and active projects

Get started free

About Nexxa.ai

Nexxa is a 2024-founded startup building AI agents that augment industrial engineers in heavy project industries. The platform targets engineers in railroads, energy, construction, and mining who operate within outdated or fragmented technology environments. Their approach centers on custom AI solutions that understand domain-specific engineering work—object detection, real-time inference, multimodal reasoning—without requiring engineers to pre-train models. Current work includes real-time inference systems, object detection and tracking, and custom generative AI solutions for manufacturing customers, with presence across US and Germany hiring.

HeadquartersSunnyvale, CA

Company Size11–50 employees

Founded2024

Hiring MarketsUnited States, Germany

Frequently Asked Questions

What AI and ML technologies does Nexxa use?

Nexxa's stack includes PyTorch, TensorFlow, TensorRT, CLIP, Vision Transformers, OpenCV, JAX, and DeepSpeed for training and inference. They also use ONNX and OpenVINO for model optimization, and Ray + Lightning for distributed ML workflows.

What is Nexxa working on?

Current projects span real-time inference systems, object detection and tracking, multimodal LLM/VLM integration, custom AI solutions for manufacturing, and large-scale ML training workflows. Focus is on generative AI and computer vision solutions for industrial engineering customers.