About the job
Join Our Team!
Become a pivotal member of the Evolv Machine Learning & Sensors team as a Senior Data Scientist. In this role, you will delve into sensor data, explore feature spaces, and ensure data quality that fuels our cutting-edge AI/ML systems. This hands-on position focuses on representation analysis, extracting exploratory data insights, and implementing data-centric enhancements that significantly elevate model accuracy, robustness, and generalization. You will navigate through classical ML and deep learning pipelines to uncover blind spots, diagnose data challenges, and shape effective data curation and collection strategies.
Your Path to Success: Key Outcomes in the First Year
First 30 Days:
- Gain a comprehensive understanding of Evolv’s sensor ecosystem, datasets, and ML pipelines.
- Analyze dataset structure, labeling processes, and existing exploratory data analyses.
- Conduct initial UMAP/PCA/t-SNE analyses to visualize data distributions and detect anomalies.
- Spot opportunities to enhance data quality, labeling consistency, and dataset coverage.
First 90 Days:
- Engage in deep representation analysis across sensor, time-series, and feature data.
- Evaluate classical ML and deep learning models by correlating model errors with data quality issues.
- Establish data quality metrics and initial dataset acceptance criteria.
- Collaborate with data collection teams to inform targeted data acquisition and relabeling efforts.
- Mine existing field data to recognize patterns and derive actionable insights.
- Devise methodologies to enhance data quality, transforming noisy or unverified data into clean, validated datasets.
End of Year Goals:
- Own data-driven insights that lead to measurable improvements in ML model performance.
- Establish continuous monitoring for data drift, blind spots, and label quality.
- Offer strategic direction for future data collection, annotation, and curation.
- Create automated tools and dashboards for data quality reporting and representation analysis.
Your Daily Responsibilities
- Data Understanding & Representation Analysis:
- Analyze high-dimensional sensor and feature data utilizing UMAP, t-SNE, PCA, and similar techniques.
- Identify clusters, outliers, distribution gaps, and blind spots within various classes and environments.
- Diagnose dataset shifts, domain mismatches, sparsity, and representation collapse.
Model-Aware Data Analysis:
- Conduct data analyses that align with both classical ML models (XGBoost, SVR, k-NN, tree-based models) and deep learning frameworks (CNNs, Transformers).
- Investigate embeddings, confusion matrices, and failure cases to connect model issues back to data causes.
Join Us at Evolv Technology
At Evolv Technology, we are dedicated to revolutionizing the way people experience spaces through intelligent sensor technology and advanced machine learning. We foster a collaborative environment where creativity and innovation thrive. Join us as we push the boundaries of technology to create safer and more efficient environments.
