The Role

As a Senior Data Scientist on Linktree’s Trust team, you will play a key role in protecting the integrity and health of our platform. You will apply your expertise in data science and statistical modeling to define how we measure harm, assess platform risk, and continuously improve our detection and mitigation systems.

In this role, you will build and scale the core measurement framework that underpins Trust at Linktree, establishing clear, defensible definitions of harm prevalence, enforcement effectiveness, and user impact. You will partner closely with Engineering, Product, Policy, Legal, and Operations to identify emerging abuse patterns, size their impact, and ensure our mitigation systems are effective and aligned with our risk tolerance.

You will also play a central role in evaluating and strengthening our ML and LLM-based safety systems. This includes defining high-quality ground truth datasets, establishing statistically sound evaluation benchmarks, and implementing monitoring frameworks to ensure model quality and reliability in production. In partnership with Engineering, you will identify failure modes, improve detection quality, and build monitoring frameworks that keep our systems effective as abuse patterns evolve.

Your work will directly shape how Linktree understands and prioritize platform risk and measures the effectiveness of its trust and safety investments.

What You’ll Own

Define and operationalize the core Trust metric frameworks that anchors company-wide north star goals and cross-functional team performance
Refine and validate the scoring model used to assess Linktree quality and risk, ensuring heuristic logic is measurable, calibrated, and aligned with policy standards
Build and maintain production-grade datasets and dashboards that provide reliable visibility into platform risk and system performance
Identify and size emerging abuse patterns through deep-dive analysis, and translate findings into clear product, policy, operations and engineering priorities
Partner with Engineering to define evaluation frameworks for ML detection and LLM-based safety systems, including ground truth design and performance benchmarking
Monitor model health in production (e.g., precision, recall, calibration, drift) and proactively surface degradation, bias, or failure model
Support Trust incident response through rapid analysis, impact sizing, and postmortem insights that drive durable improvements
Communicate risk trends and model performance insights through clear visualizations and executive-ready narratives

What You'll Bring

5+ years as a Data Scientist in a product-driven technology company
Direct experience in Trust & Safety, Integrity, Fraud, Security, or other risk-focused domains
Proven ability to define and operationalize metric frameworks (e.g., Prevalence, Recall, and False Positive Rates) to evaluate the effectiveness of product, policy, or ML-driven interventions
Experience evaluating machine learning models, including offline methodology, performance tradeoffs (precision/recall, calibration), and production monitoring
Advanced SQL proficiency and fluency in Python (or R) for advanced analytics and modeling
Strong foundation in statistics, experimentation design, and causal inference
Experience working within mature data platforms and analytics engineering stacks (e.g., DBT, Dagster, Airflow)
Experience designing intuitive, decision-oriented dashboards and compelling data visualizations using tools such as Hex, Looker, Tableau, or PowerBI
Excellent communication skills with the ability to translate complex statistical findings and risk tradeoffs into clear recommendations for cross-functional stakeholders
Bachelor’s Degree in a quantitative field (Mathematics, Statistics, Computer Science, Physics, Economics, or related); an advanced degree (MS/PhD) is a strong plus.

Linktree is committed to providing a fair and competitive compensation package including cash, equity, and benefits. Final offers depend on multiple factors including location, experience, expertise, and role scope, and may vary from the range listed.