Gal Gilor

Data Scientist

Deloitte•  February 2022 - September 2023•  Remote

As a Data Science Specialist at Deloitte, I successfully served multiple commercial clients, including Meta and Deloitte Digital, in both team lead and individual contributor roles. I adeptly handled diverse data sets and applied statistical techniques to extract valuable insights. Additionally, I demonstrated expertise in developing proof of concept models that enhanced go-to-market capabilities, resulting in tangible business success. - Decreased time to insight by 50+% through establishing and deploying automated pipelines and dashboards (Facebook internal development tool similar to Apache Airflow and Power BI) utilized by client stakeholders. - Collaborated closely with stakeholders to identify pain points in the onboarding process and developed videos and comprehensive documentation to increase knowledge transfer, resulting in a 50% reduction in onboarding time (from eight weeks to four weeks). - Enhanced team output by 300% through the creation of code templates, empowering efficient ad-hoc data pipelines, and identified and addressed three critical bugs with a global impact, resulting in notable improvements in data quality and boosting satisfaction among Meta's business partners. - Developed a focused Natural Language Processing (NLP) solution (Python, NLTK) to streamline the sorting process and reduce manual workload for government contracts within the Strategy & Transformation team—delivered a high-performance model achieving 90% lift over random chance. Leveraged SHAP and other visualizations techniques (Feature Importance, Force Plots) to enhance model explain-ability. Achieved 1st place in an internal Hackathon for the most explain-able model. - Developed and deployed an advanced question-answering application using Hugging Face and GPT-4, enhancing capabilities on the knowledge base. Implemented an efficient pipeline with Haystack and Elasticsearch to process, vectorize, and store 35,000+ client pages for seamless accessibility. - Developed an artificially generated art installation using abstractive summarization techniques (Prompt Engineering, LangChain, OpenAI, NVIDIA GPU cluster) to extract key themes around diversity, ethics, and social impact from past Cannes Festival submissions and generated captivating images using Stable Diffusion.

Data Scientist

ASR Analytics•  January 2020 - January 2022•  Remote

Led client migration from relational to NoSQL (RESTful API, Apache NiFi, MarkLogic) database, while performing data engineering (JavaScript, SQL, Python) to transform and scale solutions. - Developed end-to-end anomaly detection algorithms for tax fraud detection and provided solutions to identify fraud with 0.9 precision score (Python, Pandas, NumPy, SciPy, Scikit-Learn). - Iterated novel ensemble model approach for fraud detection, optimized auditing processes, and secured annual governmental savings >$10M. - Translated core capabilities from legacy systems to open-source technologies, while developing proprietary software and enabling ASR Analytics to include Software as a Service (SaaS) capabilities in addition to traditional IT consulting. - Championed a customer-first culture, acted as onsite customer support for all issues, and retained clients for 2021.

Tel Aviv University, Tel Aviv-Yafo

Biology, BS•  October 2014 - October 2017

Graduated Magna Cum Laude Research Focus: Insect Performance Flight and Maneuverability