Machine Learning Engineer
LeapMind is mainly engaged in the research and development of quantized neural networks and the development of Efficiera, an accelerator to run such networks at high speed, in order to solve the problem of power consumption, which is a major barrier to the widespread use of deep learning. Due to the fact that we develop our own hardware, we are looking for a machine learning engineer who is proficient in deep learning and can lead projects that utilize it in order to develop models that run at high efficiency while meeting the limitations of our hardware.
In particular, we are looking for a talented machine learning engineer who can develop highly practical deep learning models to realize edge AI that works in various industrial fields that require low-power devices, utilizing Efficiera.
Responsibilities:
・Development of quantized deep learning models considering embedded environments that run efficiently on our accelerator IP, Efficiera.
・Develop high performance, high quality models that meet business and customer requirements in target industries.
・Create product specifications, overall design, development plans, and lead design and development projects.
・Researching, understanding and sharing the latest technology trends
・You will focus on one of the following.
- Object detection
- Noise reduction for images and/or videos
- Anomaly detection
- Other applications
・Master's degree in informatics or a nearby field, or equivalent knowledge and experience.
・Experience (1+ years) leading the design and development of a product or system development project using machine learning or deep learning frameworks in a multi-person team.
・Experience (2+ years) as a developer designing, developing, and researching systems and products using deep learning.
・Deep understanding of image processing and computer vision and practical experience using it (2+ years).
・Experience in software development projects based on Agile or Scrum development (2+ years).
・Business level communication skills in Japanese.
・PhD or equivalent knowledge and experience in informatics or a neighboring field.
・Experience developing products and solutions that combine image processing and computer vision with deep learning models.
・Understanding of computer architecture and low-level programming experience.
・Experience with various model compilers and runtimes for deep learning. (TensorRT, TensorFlow Lite, OpenVINO, ONNX Runtime, etc.)
・Experience optimizing machine learning or deep learning models for embedded devices.
・Experience in accepting peer-reviewed papers on machine learning and deep learning.
・Experience with machine learning competitions (e.g. Kaggle), competitive programming. (e.g. AtCoder)
・Experience bringing products to market.
・You share LeapMind's mission and vision (https://leapmind.io/about/) and are committed to developing products that solve customer issues.
・Able to lead a company-wide project from start-up to closing, with appropriate support from others, using good judgment to lead the project autonomously to a successful conclusion.
・A team player and able to communicate with others in a respectful manner
・Able to build consensus through constructive dialogue and discussion based on trust and cooperation in order to produce optimal outputs efficiently and effectively.
・Able to extract issues from facts and events, prepare documents (proposals) that clarify goals, issues, and the best means of addressing them, and explain them to relevant parties regardless of position (we emphasize documentation when sharing information in order to ensure transparency and prevent information asymmetry)
・Share our mission and vision and take on challenges toward the same future.
Empathy with the realization of LeapMind's Mission and Vision (https://leapmind.io/about/)), interest in our technology and business, and willingness to catch up. Passionate about developing products that solve customer problems.
This is a demo video of visual inspection using the "Efficiera" anomaly detection model (Xilinx Kria KV260 Vision AI Starter Kit). Please check out the video.
https://www.youtube.com/watch?v=X4r19IbyUH8
Development language: Python, C++
OS: Linux
ML Libraries: TensorFlow, PyTorch, scikit-learn
Other: GitHub Enterprise, Slack
Company profile: https://leapmind.io/careers/
Please check speakerdeck
Employment type | Full-time |
---|---|
Minimum annual salary(JPY) | 7,000,000 |
Maximum annual salary(JPY) | 12,000,000 |
Salary description | Salary: Based on one's experience, ability, and previous salary
Bonus: Bonus to be paid twice annually based on company performance |
Location | 3F, Shibuya Dogenzaka Sky Bldg,
28-1 Maruyama-cho, Shibuya-ku, Tokyo, 150-0044, Japan (9min by walk from Shibuya Sta, 2min by walk from Shinsen Sta.) No smoking inside the building |
Apply from | Anywhere |
Remote work | Remote(Anywhere in Japan) |
Working hours | Flextime system (5:00am-10:00pm, standard working hours 8h), No core time |
Holidays | ・System of two days off a week (Sat・Sun)
・National Holidays ・Dates determined by company(New Year's holidays, etc) ・Paid vacation(5 days granted after 3 months of employment, 5 more days after 6 months of employment) ・Special holidays(Congratulation or condolence leave:Marriage, Spouse’s birth, condolence leave) ・Refreshment leave, etc. *Last year's example: New Year's holidays (12/28-1/4), Refreshment leave 4days |
Employee benefits | ・Social Insurance: Employees' Pension Insurance, Health Insurance, Employment Insurance
・Commuting Allowance: Based on the commuting route approved by the company. However, actual expenses will be reimbursed based on the number of days worked. (Maximum monthly allowance is the lower of the cost of a one-month commuter pass or 50,000 yen) ・Work remotely allowance: 5,000 yen/month ・Well-developed health management system (regular health check-ups, stress checks, Carely chat health counseling service, free flu vaccine shots, breast and cervical cancer screenings, etc.) |
Selection flow | It will approximately take 1 month from entry to job offer.
・Entry ・Document screening ・Interviews (2-3 times) ・Offer ・No dress code for the interview. Members here dress casually at work. ・We can schedule an online interview. First, please apply. ・Please refrain from re-submitting within 1 year from the previous application. ※Depending on the position, you may be required to submit a portfolio and take the skill tests. For details, please refer to the application page or the entry form for each position. |