Hybrid Solution Architect (AI/ML), Smart DC (Innovation) bei Keppel

Key Responsibilities:

Model Development

• Develop, train, and fine-tune models on time-series sensor data for both anomaly detection and failure prediction across various asset types.
• Explore and apply a mix of forecasting, anomaly detection, change point detection, survival analysis, and representation learning techniques, choosing the best ft based on the use case.
• Be comfortable applying hybrid approaches (e.g., combining statistical thresholds with ML models, or chaining changepoint detection with LSTM/Transformer predictors) when appropriate. Iterate quickly and deliver working ML models into dev or production environments every 1–2 week sprint, enabling rapid feedback and continuous improvement.
• Balance performance, explainability, compute cost, and deployment constraints in every modeling decision

Stakeholder Communication

• Clearly explain model behavior (e.g. thresholds, decision boundaries).
• Share experiment results, performance comparisons, and trade-offs transparently.
• Collaborate with product managers, engineers, and non-technical stakeholders.
• Align ML decisions across the team for unifed communication.

Performance Monitoring

• Continuously monitor model performance in production (e.g., accuracy, drift, recall).
• Build a retraining and rollback strategy to handle data drift, model drift or edge cases.
• Use dashboards or alerts to track live model degradation.
• Proactively recommend model retirement or replacement.

Platform Enhancement

• Suggest improvements to model architecture, data input, and system-level optimization
• Balance performance and compute/resource efficiency
• Contribute to architecture discussions with an awareness of cost, scalability, and maintainability

MLOps & Pipeline Collaboration

• Work closely with MLOps engineers to build versioned, reproducible training/inference pipelines.
• Provide clear model packaging requirements (inputs, outputs, dependencies).
• Understand CI/CD for models and how to deploy across dev, test, and prod.
• Support auto-retraining and rollback pipelines.

Data Integration

• Work with engineers to integrate real IoT data (modbus, BACnet, analog/digital sensors).
• Understand data latency, time sync issues, missing values, and signal quality.
• Find the right balance between capturing enough sensor detail for useful signal processing and avoiding excessive noise or redundant data
• Advocate for a balanced IoT pipeline design — leverage edge intelligence where appropriate (e.g., thresholding, compression, metadata tagging) instead of relying on "dumb edge" devices that push raw data continuously.
• Prioritize efficient data processing and storage — enabling scalable, cost-effective ML workflows.

Qualifications

• Bachelor’s or Master’s degree in Computer Science/Statistics/ Data Science, or related field.
• Proven experience building, deploying, and monitoring time-series ML models.
• Strong Python skills, with experience in TensorFlow, PyTorch, or Scikit-learn.
• Experience with time-series modeling (LSTM, GRU, TCN, attention-based).
• Strong understanding of Generative AI: prompt engineering, fne-tuning, foundation models.
• Knowledge of transfer learning for adapting foundation models to predictive maintenance.
• Familiarity with Azure ecosystem: IoT SDK, Event Hub, Databricks, Azure ML.
• Experience working with analog/digital sensor data, and real-world data pipelines.
• Strong communication and collaboration skills across product and engineering teams.
• Experience in working under Engineering environment