Junior Data Scientist

The Role

You’ll support Apptegy’s personalization initiatives by using data and applied modeling to make the parent experience more relevant, timely, and engaging. Day to day, you’ll help develop, evaluate, and iterate on the models and algorithms that turn real-world school communications into actionable signals and personalized experiences.

You’ll use product and engagement data to engineer features, assess model performance, and measure impact through experiments and feature readouts. You’ll also help define what “quality” means for these systems—building test sets, partnering on annotation, and tracking accuracy over time—so we can improve personalization confidently and responsibly.

In partnership with Data Science, Product, Engineering, and Design, you’ll translate messy, unstructured language data and behavioral insights into clear recommendations and product improvements. This is a great role for someone who enjoys both analytical work and cross-functional collaboration, and who also likes getting into the weeds of unstructured communications to extract what is most important and actionable.

What You’ll Do

Modeling & Reporting

• Build and evaluate lightweight models/scoring signals (e.g., likelihood to engage, propensity, relevance) using sound feature engineering and performance metrics.
• Create repeatable reporting (dashboards, recurring analyses, experiment readouts) to monitor personalization performance over time.
• Develop and use ETL pipelines that transform raw text communications into structured datasets that enable analysis, reporting, and insight discovery.

Analysis & Insight Discovery

• Analyze parent engagement and product behavior using funnels, cohorts, segmentation, and journey analysis.
• Define and track personalization success metrics (engagement, retention, adoption, satisfaction proxies).
• Build and maintain representative evaluation datasets (“test sets”) and help support annotation workflows so we can measure accuracy and consistency over time.
• Deliver concise insight readouts that recommend what to personalize, for whom, and what to test next, including where performance varies by context (elementary vs junior high vs high school; operations vs classroom vs athletics/clubs).

Collaboration

• Partner with Product, Design, and Engineering to shape measurement plans and evaluate experiments or feature rollouts.
• Help validate instrumentation, resolve data gaps, and document metric definitions and methods so reporting is trusted and repeatable.
• Contribute to a quality and accuracy culture by sharing learnings from evaluation and error analysis in a way that helps the team iterate quickly.

Who You Are

• You have 1–3 years of experience in data science, analytics, or a related role.
• You’re strong in SQL and comfortable in Python (pandas/numpy), with working knowledge of statistics and experimentation.
• You enjoy insight discovery—digging into messy product data, spotting patterns, and turning them into clear takeaways and next steps.
• You can build and evaluate lightweight models/scoring signals and you care about doing it responsibly (sound evaluation, monitoring, and iteration).
• You communicate clearly with both technical and non-technical partners, and you collaborate well across Product, Design, and Engineering.

What Makes You Stand Out

• Hands-on experience with a modern data warehouse (Snowflake/BigQuery/Redshift/Databricks) and writing efficient SQL on large event datasets.
• Familiarity with data build tools or analytics engineering practices (tested models, documented metrics, data quality checks).
• Applied ML experience in Python (scikit-learn or similar): feature engineering, model evaluation, and basic monitoring.
• Exposure to personalization/recommendation concepts (segmentation, propensity scoring, ranking signals) and how to measure impact.
• Experience working with or evaluating NLP/LLM systems, especially around accuracy measurement, dataset construction, and error analysis.