Datakrew is a deep-tech startup based in Singapore, focused on building next-generation analytics for electric vehicle (EV) fleets, batteries, and grid-connected energy assets.
The OXRED platform is deployed across OEMs and operators to extract real-time intelligence from EV battery telemetry, reduce safety risks, and extend usable battery life.
The company is looking for a Battery Intelligence Engineer to support internal R&D and analytics module development.
Responsibilities include building, training, and benchmarking ML models for battery State-of-Health (SOH), Remaining Useful Life (RUL), and degradation prediction.
The role involves analyzing large-scale battery performance data to extract actionable insights and developing anomaly and event detection algorithms for safety-critical battery threats.
The engineer will create predictive maintenance models to forecast battery failures and degradation, validate and benchmark model performance using real-world and simulated datasets, and work with the product team to ensure integration of developed models.
Staying updated with the latest research in battery modeling and machine learning techniques is essential, as well as researching and collaborating with the team to incorporate electrochemical knowledge into ML models.
Requirements:
Proficiency in Python, including libraries such as Pandas, NumPy, Scikit-learn, and either PyTorch or TensorFlow is required.
A strong grasp of time-series modeling, signal processing, and statistical evaluation is necessary.
Familiarity with LSTM, GRU, transformer models, or forecasting pipelines is preferred.
Hands-on experience with deep learning models (CNN, DNN, RNN), NLP techniques, and sequence models is essential.
Exposure to modern LLM ecosystems (ChatGPT, Gemini, Ollama) and tools like LangChain, AI agents, and prompt engineering is beneficial.
A good understanding of battery systems, including SOC, SOH, DOD, thermal behavior, and lifecycle stressors is required.
Experience working with EV, IoT, or BMS datasets, even in academic projects, is necessary.
Basic SQL knowledge for querying historical telemetry is required.
Familiarity with Kalman filters, physics-informed models, or hybrid ML-physics methods is a nice-to-have.
A GitHub portfolio or academic project code repositories are advantageous.
Knowledge of EV architectures or energy storage standards is also a plus.
Benefits:
The position offers flexible remote work arrangements.
The engagement duration is between 2 to 4 months, with the possibility of extension.
This is a part-time or freelance position, providing opportunities for individuals interested in the EV or battery analytics domain.
