Ver categorias
Explorar
Fiverr Pro
Português
$
USD
Hazeldonaldson
@hdonaldson_95
Data Science based account
Estados Unidos
Sobre mim
Senior data professional (5 years) building predictive models and NLP/LLM systems for subscription, health tech and EdTech companies. Predictive modeling: churn, fraud, demand forecasting, and risk scoring with XGBoost and scikit-learn. NLP/LLM: text classification with Anthropic APIs, BERT, and vector databases. I work end-to-end - data cleaning, feature engineering, model selection, and validation- delivering a script, API, dashboard, or recommendation you can act on, not just a notebook. Recent work: a review-rating text classification using TF-IDF.
Habilidades
Hazeldonaldson
offline •
Conheça meus serviços
Machine learning
I will create a custom machine learning model for you business data
Experiência profissional
Data Analyst, Growth
SonderMind • Período integral
Jan 2025 - Present • 1 yr 9 mos
* Integrated 5+ new data sources to expand Lifecycle tracking capabilities, enabling more comprehensive customer journey analysis and improving the Marketing team’s ability to target key conversion points with 15% higher precision * Built client personas for the Marketing team by analyzing behavioral and demographic data across 10,000+ active clients, resulting in more targeted campaign strategies and a 12% improvement in campaign engagement rates * Partnered with the Product team to design and implement tracking infrastructure for a new client portal, delivering critical behavioral and usage data that informed UX decisions and supported a successful portal launch to 8,000+ clients * Developed a new active client forecast model to predict monthly client acquisition, achieving 85% forecast accuracy and enabling more effective resource allocation and growth planning.
Data Scientist
CommonLit
Feb 2022 - Jan 2025 • 2 yrs 11 mos
* Collaborate with a cross-functional team of data scientists, researchers, and product managers to analyze complex data sets and user data, contributing to a 20% improvement in student reading scores across three partner districts. * Extract, clean, and analyze data sets using various data analysis tools such as Looker, Python, and R, identifying patterns and trends to optimize performance and drive a 20% increase in student reading proficiency. * Conduct 30+ ad-hoc analyses per quarter on application and event stream data streams, enhancing data accuracy by 20% and supporting key decision-making processes by communicating findings to non-technical stakeholders. * Maintain over 100 dashboards in Looker, reducing time spent on routine data requests by 30%, achieving about $150,000 in cost savings per quarter, and improving report accessibility by gathering and integrating stakeholder data requirements. * Develop 100+ unit tests in LookML to boost data integrity by 25%, ensuring consistent, high-quality data insights. * Partner with product, sales, fundraising, and marketing teams to align data insights with strategic objectives, delivering actionable recommendations to stakeholders based on business metrics, increasing decision-making efficiency by 15%.
