Job Description
We are looking for a Data Scientist to join our client's R&D team, a small group of scientists and engineers developing algorithms that help improve the health of millions around the globe. You will design, develop, and optimize algorithms on real-world multi-channel sensor data, working directly with the scientists who own the research. We are looking for a candidate with a strong theoretical foundation in the mathematics behind algorithm development, a solid background in mathematics, physics, electrical/biomedical engineering, or computer science, recent industry experience with sensor signals, and a passion for solving complex problems.
Responsibilities
Design, implement, and optimize algorithms for data processing, analysis, and decision-making systems on multi-channel sensor data.
Process raw sensor signals: filtering, denoising, windowing, FFT/spectrogram analysis, and feature extraction from raw signals.
Combine multiple sensor channels recorded in the same time window into a single model.
Translate theoretical concepts into production-quality Python code.
Identify, assess, and account for confounding factors in data analysis and model validation.
Own the ML pipeline end to end, from raw data to validated model, including training on Google Cloud.
Analyze algorithm performance, complexity, and numerical stability.
Document algorithm design, assumptions, and limitations, and communicate results directly with the client's scientists.
Requirements
At least 4 years of experience.
Bachelor's degree (BSc) in Mathematics, Computer Science, Electrical/Biomedical Engineering, Physics, or a related field.
Strong mathematical background and the ability to independently take a problem from concept to implementation.
Solid programming skills in Python.
Recent hands-on experience with multi-channel time-series (e.g., 20+ sensor channels in the same time window) combined into a single model, with at least 2 projects in the last 24 months.
Practical signal processing on raw sensor data: filtering, denoising, windowing, FFT, spectrograms, feature extraction, and unsupervised feature selection.
Solid understanding of confounding factors and proven use in real projects, including the ability to identify, assess, and account for them in data analysis and ML models.
Experience with both CNN and non-CNN architectures, including LSTM, for sequential data.
End-to-end ownership of an ML pipeline, from raw data to validated model, including training in cloud environments (Google Cloud preferred).
Industry experience with large, noisy, real-world data (academic-only experience is not sufficient).
Nice to have: Experience working in a small R&D team (under -15 people) or startup, collaborating directly with scientists.
Nice to have: Master's degree (or currently pursuing) in a relevant field.
Nice to have: Experience with biosignals or sensor-array data.
Nice to have: Familiarity with algorithm optimization for performance and scalability.
Nice to have: Understanding of edge-device deployment (deployment itself is owned by MLOps).
Why you'll love working at Eastgate
Attractive salary package.
Eastgate Software is a Global Fortune 500 Preferred Vendor, so you will have an opportunity to work with many IT experts around Europe and USA.
Be emerged in a professional and motivating environment where you have opportunities to express your own ideas and improve your skill set through many interesting full-cycle projects.
Flexible check-in: 8:30 - 10:30AM, Monday to Friday, No Overtime (Only 8 working hours).
Performance review and appropriate gratification: twice a year.
13th month salary + Project bonus + 02 probationary months with full salary.
Extra health check package per year.
16 leave days/year (with 12 annual leave days and 4 paid leave days for learning activities). You will be added an extra day each year and a maximum of 20 days for annual leave.
Free snacks, video gaming & movie nights, team social and sport activities.
Thông tin chung