H

H Company Tech Stack

AI research and implementation platform for enterprise LLM and agent deployment

Technology, Information and Internet Paris, Île-de-France 51–200 employees Privately Held

H Company operates a machine learning infrastructure and research outfit built on PyTorch, JAX, TensorFlow, and a deep optimization stack (CUDA, Triton, vLLM, TensorRT-LLM). The tech foundation—spanning GPU kernels, inference pipelines, and training infrastructure—reveals a company shipping inference and training systems to enterprises, not just consulting. Hiring skews research and engineering (14 of 22 roles) with aggressive velocity, signaling scaled delivery of proprietary models and agent platforms.

Tech Stack 51 technologies

Core StackPython PyTorch TensorFlow Rust Go Node.js Java React Vue Flask Django FastAPI AWS GraphQL RabbitMQ Kafka Git JAX C/C++ CUDA Triton vLLM TensorRT-LLM SGLang llama.cpp NCCL ONNX Runtime ExecuTorch REST AWS SQS+21 more

What H Company Is Building

◆Challenges

Optimizing hardware utilization
High throughput and low latency
Soc 2 compliance
Ensuring gdpr compliance
Implementing agents at scale
Integrating with customer systems
Iterating product fit
Recruiting elite talent
Optimizing recruitment processes
Managing coworking space logistics

▲Active Projects

Training infrastructure for model and agent development
Llm and vlm training
Scalable inference pipelines
Agent platform apis and runtimes
Observability and monitoring
Enterprise platform development
Observability suite development
Gpu kernel development
Instruction following and tool use research
Model performance optimization

Hiring Activity

Accelerating20 roles · 10 in 30d

Department

Engineering

9

Research

5

Marketing

2

Ops

2

Product

1

Sales

1

Security

1

Seniority

Senior

8

Mid

3

Staff

3

Manager

2

C-Level

1

Director

1

Intern

1

Lead

1

Notable leadership hires: Sales Director

Company intelligence

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

Get started free

About H Company

H Company develops AI research and deployment infrastructure for enterprise customers. The product suite spans training infrastructure for custom models and agents, scalable inference pipelines, and an agent platform with APIs and runtimes. The tech stack shows deep systems work: GPU kernel development, observability and monitoring pipelines, and performance optimization across the inference chain. The company is based in Paris and actively recruiting across France, the UK, and the US.

HeadquartersParis, Île-de-France

Company Size51–200 employees

Hiring MarketsFrance, United Kingdom, United States

Frequently Asked Questions

What tech stack does H Company use for LLM training and inference?

H Company uses PyTorch, JAX, TensorFlow for training; Triton, vLLM, TensorRT-LLM, and SGLang for inference optimization; and CUDA, NCCL, and custom GPU kernels for hardware acceleration. FastAPI and REST serve the API layer.

What is H Company building?

H Company is building training infrastructure for models and agents, scalable inference pipelines, agent platform APIs and runtimes, and GPU kernel optimizations. Recent focus areas include observability suites, instruction-following research, and enterprise platform integration.

Similar Companies in Technology, Information and Internet

Other companies in the same industry, closest in size

R

Ruby Labs

Technology, Information and Internet

C

Candidata ai

Technology, Information and Internet

K

Kanz

Technology, Information and Internet