Avalanche Computing Taiwan Inc. cover image
Logo of Avalanche Computing Taiwan Inc..
Logo of Avalanche Computing Taiwan Inc..
Avalanche Computing Taiwan Inc.
Logo of Avalanche Computing Taiwan Inc..
Avalanche Computing Taiwan Inc.
Logo of Avalanche Computing Taiwan Inc..
Avalanche Computing Taiwan Inc.
About

Company summary

YouTube video player

Avalanche Computing Taiwan Inc. was funded by Nvidia USA alumni in Santa Clara, CA, USA, in 2018. The core team was established in Taipei, Taiwan at the end of 2018. After a year of collaboration with many enterprises, Avalanche Computing Taiwan Inc. was registered in February 2020. In the first month, we won the Airbus global top-10 innovation startup award in March, joined the NVIDIA global inception program, and became a part of NTUTEC. The time flies, we won several awards in global huge exhibitions at Viva Tech 2023, LTW 2022, QITC 2021, et., al.

What we do:
The barrier to entry in AI is lower than ever before, thanks to open-source software, including frameworks like TensorFlow, Keras, PyTorch, and more. However, developing specific AI applications still poses challenges for traditional companies and SMEs, as they must build data pipelines, computing environments, and AI models. But, only a few teams can help speed up the AI computing (NVIDIA team and we can do well).

To overcome these challenges (few talents, long-time waiting for computing, and a long-development cycle), we provide hyper-scale computing techniques for deep learning. Our hyper-scale computing framework is a deep-tech solution for computing technology. With our hyper-scale technology, clients can focus on innovation while we handle the rest. We offer hyper-scale architecture, multi-GPUs, distributed model training, and hyper-scale inference on 10,000+ edge devices. Clients can simply upload their AI models into our SaaS AI development tool within minutes. We call it hAIsten AI.

We also won several awards:

  • 2024 SusHi Tech Tokyo (May 15-16 the selected startup by Japan gov)
  • 2024 GTC San Jose (March 18-21 Feature startup and booth with Taiwan partners)
  • 2024 CES per-roadshow, best investment startup in the Bay Area (Jan 5, 2024) https://www.ntdtv.com/b5/2024/01/25/a103846335.html
  • 2023 SLINGSHOT by SWITCH Top-50 (OCT 6 updated)
  • 2023 Viva Tech Startup Pitch Award (Edge Hub )
  • 2023 Apr, SelectUSA Taiwan startup delegate
  • 2022 Dec, TAcc+ Winner for Global Expansion 
  • 2022 Intel IPA (member), WPI (business partner)
  • 2022 Featured startup in London Tech Week
    (https://www.bnext.com.tw/article/69944/most-tta-startup-uk
  • 2021 Best Investment Startup in APAC (EETimes)
    https://site.eettaiwan.com/emag/202112/files/basic-html/page9.html
  • 2021 Best Investment Startup in Taiwan (EETimes)
  • 2021 QITC 10 Finalists (Qualcomm)
  • 2021 Top 200 teams (South Summit)
  • 2021 Top DeepTech Pioneer (HelloTomorrow Global 2021)
  • 2021 Berkeley SkyDeck GIP
  • 2020 NVIDIA Inception
  • 2020 NTUTEC alumni
  • 2020 The global finalists by Airbus (HelloTomorrow)

Important Expos:

  • NVIDIA GTC San Jose (2024)
  • Techcrunch (2020, 2021)
  • CES (2020, 2021, 2022, 2023, 2024)
  • VIVE Tech (2022, 2023, 2024)
  • London Tech Week (2022, 2023, 2024)
  • Hub Berlin! (2022)
  • Collision (2021)
  • HelloTomorrow (2020, 2021)
  • South Summit (2021, 2022)
  • Websummit (2021, 2022, 2023, 2024)
  • AI Summit (2022, 2023)

Our office is now in the following

  • Taipei (office), Taiwan
  • San Francisco (HQ moved 2025), CA, U.S.
  • Berlin, GE (Coming soon)
  • London, UK (Coming soon)
  • Nagoya and Tokyo, Japan (Coming soon)
  • Paris, FR (Coming soon)

Now, the SaaS product is launched in the U.S. 

If you are interested in our position, please send your CV to: [email protected]

Products or services

We provide performance-optimized AI development and deployment tool (hAIsten AI) to data scientists in a small AI team. It is a low-code tool to speed up your AI development with clicks-through. Our core technique, Hyper-scale computing tech, leverages the distributed and multip-GPUs techniques for saving time and money for your AI development (time-to-market).

The product of Avalanche Computing is called hAIsten AI in the US. The detailed functions are as follows:

Phase 1 - hAIsten AI

- Hyper-scale training computing engine ( >1 GPU on each machine)

- Hyper-scale inference engine on the cloud (AWS, Azure, and GCP)

Phase 2 - Edge Hub

- Edge computing platform (CES 2023 launched)

- Deploy your AI to all Edge devices in seconds (from 1 edge device to 100000+ edge devices)

- No code for AI conversion and optimization

Other tools (it may be from partners)

- Smart labelling (AI-empowered data annotation tool)

Mission

Vision 

is to be the best advanced and hyper-scale computing technology company for intelligence application providers and traditional industries, and in collaboration with our clients, to build a strong competitive power in the intelligence application industry.
To achieve our vision, we must have the following advantages:
1) be a computing technology leader, competitive with pure hardware or software solutions
2) be the top-tier computing technology, data analysis, solution architector, and AI experts
3) be the most trusted, service-oriented, and maximum-total-values in intelligence development software end-to-end solution inventor

Mission 

is to provide the best AI development tool and be the leader, and trusted computing technology provider in the intelligence application industry (for medical, manufacturing, et., al) for the future.

Employee benefits

We offer (depending on team type, funding team, core team, or full-time employee):

  • Competitive salary and extensive social benefits
  • Work-life balance and support for career development
  • Learning resources for online training and physical training (including conference, meetup, book)
  • Snack bar, refrigerator, snack, coffee machine with beans, red tea, japan tea, green tea, US flower tea, or milk tea from Japan (日東紅茶)
  • Company outings, festive parties, lunch days, tail teeth, employee travel (2019 we went to Japan), birthday cake
  • Company travel
  • Visa assistance if needed (Need pass - 6 months qualification exam)

Work environment

Diverse and dynamic work environment (smart office)

  • Work at home or the office (US)
  • Work at the office in Taipei (TW)
  • Work at the co-working space in Hsin-Chu City (TW), Tokyo and Nagoya (JP)
  • Work at home (UK, GE, FR)
  • Work remotely during a business trip (Global)
Avatar of the user.
AI/Deep Learning Specialist

Tech stack

webpack
Webpack
yarn
Yarn
vuejs
Vue.js
ssh
SSH
slack
Slack
postgresql
PostgreSQL
javascript
JavaScript
figma
Figma
docker
Docker
cloudflare
Cloudflare
amazonwebservices
Amazon Web Services
azure
Azure
scala
Scala
Distributed Algorithm
Distributed Algorithm
cplusplus
C++
confluence
Confluence
Flask
Flask
NVIDIA TensorRT
NVIDIA TensorRT
linux
Linux
Parallel Programming
Parallel Programming
Machine Learning
Machine Learning
Data Mining
Data Mining
Others
Others
Deep Learning (NLP / Computer Vision)
Deep Learning (NLP / Computer Vision)
java
Java
gitlab
GitLab
confluence
Confluence
JIRA
JIRA
github
GitHub
kubernetes
Kubernetes
NVIDIA Data Center Solution
NVIDIA Data Center Solution
python
Python
Tensorflow-lite
Tensorflow-lite
docker
Docker
PyTorch
PyTorch
tensorflow
TensorFlow

FAQ

是否提供工作簽證?

無,本公司為新創公司,不利於協助您申請台灣工作簽證。

是否需要加班?

視情況而定,原則上按照法定工時,每日8小時,每週40小時為限。
本公司文化延續在矽谷工作時之習慣,不做Micro Management,加班需要有理由,按照工作緩急輕重應該作個人化管理。
請務必,盡量在時間內完成工作,若是無法達到工作進度則表示個人能力問題,不期望以加班方法解決。

註: OKR 與KPI 並行,須自我不斷精進。 

奎景運算科技股份有限公司 (Avalanche Computing Taiwan)的團隊文化是什麼?

我們的公司文化有10個Key,可實際在面談中感覺你是否符合10個重要的文化特質。

公司十分注重對客戶的誠信價值,個人的主動積極態度,是否是團隊工作者。

(我們不期望因為要符合面談,而強加去更動您的履歷設計。我們想看到的是最真實的您。)

上下班時間為何?

原則上:

每日上午 9:30-10:30之間到公司或者remote work,按台灣公司勞基法準則,需要做打卡的一個動作 (使用軟體)。
每日工時以 8小時為基準。例如:9:30 AM 抵遠公司,則在7:00 PM (中間午休1.5小時)。

實務上,Agile的工作習慣以及責任制的精神 (同矽谷時期之工作經驗)。

加入 奎景運算科技股份有限公司 (Avalanche Computing Taiwan) 可以得到或學習到什麼?

無法自公司背動學習任何技能, 本公司的的人員之技能與許多知識採用自我學習,若工作需求上需要可以申請課程補助、活動補助。
至多,實質上獲得是薪水,學習到團隊合作以及在新創公司的待人處事。
公司也有訂閱線上課程以及核心團隊己取得之授課資源 (例如:GCP USA, NVIDIA Santa Clara 之課程資源)

奎景運算科技股份有限公司 (Avalanche Computing Taiwan) 的近期和長期發展分別是什麼?

本公司近期建立台灣之研發團隊,中長期返美持續拓展 Hyper-Scale Computing 之最佳化技術予服務智慧應用之廠商。 

  • 2021年, 己返美成立美國團隊並作為主力市場發展
  • 2023年, 開始拓展歐洲市場並重啟日本市場

投遞履歷時,是否有什麼注意事項?

  • For all candidates
    本公司為新創公司,是否能符合早期公司文化與需求為主要依據。現因人手問題,我們僅主動通知符合資格的求職人員,若是不符合未能及時回覆,敬請原諒。 
    Our company is a startup, and we mainly focus on early-stage company culture and requirements. Due to manpower limitations, we only proactively notify qualified job applicants. If you don't receive a timely response, please accept our apologies.
  • 請展現真實的自己即可

刷題或者競賽可以增進您的實力,若沒有長期閱讀論文、研究、自我要求的習慣,一樣是不會有效,所以請展現真實的您。(我們面對的問題都是目前國際上難解的難題,沒有參考解答) 

  • 台灣 Intern 求職事項

我們需要可以長期任職 (1年以上 Intern時間),至少完成完整的 Rotate Program 設計的人選,短期的人員或者超過資格 (Over qualify) 的人員我們雖想合作,但仍然以可以完整走完Rotate Program的人選為優先,作為長期培育之人才投資。碩博士級學生優先,需指導教授同意(我們一定會私下找您的老闆/教授聊聊,請務必和教授先作好生涯之規劃)。

  • For International Students (Interns)

The position in cakeresume is for TAIWAN ONLY. Do not submit for the position Here (go to LinkedIn).
Before you apply for any position here, please check your location and VISA status; we cannot provide you with a working permit in Taiwan (before you pass the test and work with us for more than 6 months). 

Unpaid- Intern (Taiwan local) Avalanche Computing, a deep learning data scientist / engineer, delivers amazing quality models and algorithms by combining state-of-the-art computer vision, image pro...
台灣臺北
Regular earnings reach NT$40,000
No requirement for relevant working experience
No management responsibility

Research and evaluate different AI and Deep learning approaches to determine the most effective solutions for given problems Design AI and LLM system pipelines, leveraging libraries like PyTorch, S...
台灣台北市
35.75K ~ 89.7K TWD / month
1 years of experience required
No management responsibility

Data Scientists (AI) Intern (Taiwan local) Avalanche Computing deep learning data scientists deliver amazing quality models and algorithms by combining state-of-the-art computer vision, image proce...
台灣臺北
183 ~ 240 TWD / hour
No requirement for relevant working experience
No management responsibility