S

SenseTime 商汤科技 Tech Stack

Multimodal AI platform spanning vision, language, and generative models

IT Services and IT Consulting Hong Kong, Hong Kong 5,001–10,000 employees Founded 2014 Public Company

SenseTime is a public AI software company building generative, vision, and decision-intelligence models at scale. The tech stack reflects a mature ML infrastructure operation: TensorFlow, PyTorch, MLflow, and Langchain for model development; Docker and Kubernetes for orchestration; and NVIDIA hardware (including Jetson edge devices) for inference. Active projects range from pretraining large language models to embedded systems for robotics, signaling diversified commercialization across both cloud and edge deployments. Hiring has decelerated but remains concentrated in engineering and product roles, with a notable pipeline of interns—typical for research-led AI orgs scaling from academic roots.

Tech Stack 30 technologies

Core StackAfter Effects Python Go Rust TensorFlow PyTorch MLflow Langchain RAG Linux C++ Solidworks Docker Kubernetes Photoshop C/C++ Git CAN UDP TCP Horizon NVIDIA STM32 NVIDIA Jetson C Excel Word WeChat Microsoft Office Bash

What SenseTime 商汤科技 Is Building

◆Challenges

Meeting sales targets
Reducing manual errors

▲Active Projects

Ai platform project
Ai medical product database building
Quadruped robot dog embedded system integration
Offline event execution
Machine learning development ecosystem
Deep learning software stack optimization
Ta talent management project
New media operation
Official public account operation
Pretraining language models

Hiring Activity

Decelerating40 roles · 6 in 30d

Department

Engineering

9

Product

6

Marketing

5

Sales

4

HR

3

Research

3

Data

2

Support

2

Seniority

Intern

27

Manager

4

Mid

3

Junior

1

Senior

1

Company intelligence

Find more companies like SenseTime 商汤科技 by tech stack, pain points and active projects

Get started free

About SenseTime 商汤科技

Founded in 2014 and publicly traded, SenseTime operates as an AI software platform company with 5,001–10,000 employees headquartered in Hong Kong. The company's technical focus spans multimodal model research (vision, natural language processing, decision intelligence) and infrastructure work in AI chips, sensors, and computing systems. Projects under way include medical AI databases, quadruped robot embedded integration, and large language model pretraining. The organization sells into enterprise and automotive verticals; they also maintain standards-development and governance activities in AI ethics. Sales and marketing functions exist but are smaller than engineering—consistent with a research-first business model.

HeadquartersHong Kong, Hong Kong

Company Size5,001–10,000 employees

Founded2014

Hiring MarketsChina, Thailand, Singapore

Frequently Asked Questions

What AI models and frameworks does SenseTime develop with?

TensorFlow, PyTorch, and MLflow for model training and lifecycle management; Langchain and RAG (Retrieval-Augmented Generation) for language applications. Infrastructure runs on Docker, Kubernetes, and NVIDIA GPUs including Jetson edge devices.

What projects is SenseTime currently working on?

Active projects include pretraining large language models, medical AI product databases, quadruped robot embedded system integration, deep learning software stack optimization, and machine learning development ecosystems.