DATA SCIENTIST

Experience: 7+ Years

Location: Remote

Duration: 3 Months+ Extendable

KEY RESPONSIBILITIES

1. Data Processing & Signal Analysis

· Process raw gravity, magnetic, IMU, and GNSS sensor data.

· Develop filtering, denoising, drift correction, vibration correction, and resampling pipelines.

· Convert time-series sensor outputs into calibrated spatial anomaly profiles.

2. Simulation & Synthetic Data Development

· Generate synthetic datasets from physics-based forward models.

· Add realistic noise (instrument, environmental, vibration) to training data.

· Create labelled hazard-detection datasets covering voids, water ingress, weak zones, utilities, etc.

3. Machine Learning & Anomaly Detection

· Design, train, and evaluate models for:

- anomaly detection

- classification of subsurface hazards

- uncertainty quantification

· Use classical ML (Random Forest, XGBoost, SVM) and deep learning (CNNs, LSTMs, autoencoders, transformers).

· Develop semi-supervised and unsupervised models for scarce labelled data.

4. Multi-Sensor Data Fusion

· Build fusion architectures that combine:

- Data A

- Data B

- Data C

- Data D

· Implement early, mid-level, and late fusion strategies.

· Estimate anomaly probability scores with confidence intervals.

5. Spatial Analytics & Geospatial Modelling

· Convert anomaly scores into geospatial hazard maps.

· Use GIS tools (QGIS/ArcGIS) for spatial interpolation and corridor mapping.

· Assist in building 1D/2D/3D visualisations for end-users.

6. Technical Reporting & Demonstration

· Document all pipelines, algorithms, and findings.

· Support feasibility reporting business case.

ESSENTIAL SKILLS & EXPERIENCE

· AI / Machine Learning

· Strong expertise in Python (NumPy, SciPy, Pandas, PyTorch and/or TensorFlow).

· Experience with time-series modelling, anomaly detection, and deep learning architectures.

· Experience designing ML pipelines from scratch.

· Signal Processing

· Proficiency in filtering, PSD analysis, spectral methods, and Kalman filtering.

· Experience handling noisy, multi-sensor, real-world measurement data.

· Geospatial / Modelling

· Ability to work with spatial data (GIS, coordinate transforms, DEMs).

· Experience with geophysical modelling or physical-simulation-driven ML (advantage).

· Software Engineering

· Ability to write clean, reproducible, version-controlled scientific code.

· Experience building modular ML pipelines.

· Communication

· Ability to write technical documentation clearly.

Job Type: Contractual / Temporary
Contract length: 3 months

Work Location: Remote