Senior Data Engineer

Data Engineer

Uplift Inc.

TasteGuru.ai

• Conducted competitor analysis and identified the niche and target clients for the product

• Established a text data processing ETL in PySpark framework to clean, extract noun phrases, categorize and visualize the flavor distribution from raw twitter data

• Built sentiment analysis models based on Yelp review data using flair (PyTorch based) with text embedding trained by ELMO, Vader and TextBlob.

• Designed and deployed algorithms into EMR (Spark service in AWS) to measure the novelty and trend score of different CPG products.

Data Scientist

• Anomaly detection: Built machine learning models using XGBoost, auto encoder to identify fraud population in different customer segments and saved additional 13% of the fraud loss amount in the entire channel.

• Data engineering : designed and built tables, aggregating multiple data sources (Web, SQL, flat files), as part of a greater initiative to build a new data warehouse. (mssql)

• Operations Research: redesigned and finalized a case assignment algorithm implementing a version of Dijkstra(Maximum flow) to assign cases based off analysts' skills and capacity as well as urgency of cases as a greater initiative of maximizing throughput and ensuring optimal scheduling of tasks

• Software development and ETL : Automated 3 major BAU reporting systems for fraud Claims department and deployed the program in production by constructing scripts that compiled and scheduled data from multiple sources (SQL, Web, online shared drives). The automation saved 1 FTE for the Claims department. (Python, Luigi, web scraping, VBA)

• Leadership : Developed leadership skills by creating a project plan following the CRISP-DM model and holding weekly scrum meetings to ensure project deadlines were met in a timely manner

•	Natural Language Processing: Designed an algorithm that dynamically generates tags for 244 cyber attacks by incorporating topic modeling, clustering and word embedding. Achieved 83% precision and 80% recall rate on the test dataset.

•	Log Analysis: Constructed a rule synthesis function for Splunk Enterprise to automatically detect malicious data based on advanced probability analysis and anomaly detection on metadata. Achieved an overall ROC score of 85% on the test dataset

•	Software development: Adapted GitHub/GitLab for version control and modulated algorithms into agnostic packages to process batch data in production

Data Scientist associate Intern

Accenture

•	Conducted market research to identify indicators and standards for parcel damage from the rivalry companies and proposed the preliminary strategy plan for the clients.

•	Brainstormed the root cause analysis prone to parcel damage to classify the problem types

•	Performed data analysis using Python and Excel to identify major operation procedures that cause parcel damage

•      Identified the major parcel damage transit centers by designing on-site experiments to investigate detailed operations

•	Designed part of the final project report slides

Data Analyst

Boehringer Ingelheim

•	Translated technical materials and operation manuals
•	Participated in basic quality control labs and record data
•	Completed supplier qualification
•	Developed teambuilding project plans

Summer Intern

Covestro

•	Translated technical materials and operation manuals
•	Tested intermediate product and record the data
•	Collated and correct P&ID 
•	Learned the structure of valves and pumps
•	Conducted the regular weekly check to the production plant followed by the supervisor’s instructions
•	Compiled and assorted technical files

Taplio

Priscilla Zhang's Linkedin Analytics

Want to drive more opportunities from LinkedIn?

Taplio

.css-1jgreyo{background-image:linear-gradient(284deg, rgba(39,43,72,1) 0%, rgba(33,47,183,1) 100%);color:transparent;-webkit-background-clip:text;background-clip:text;}Priscilla Zhang's Linkedin Analytics

Want to drive more opportunities from LinkedIn?

Priscilla Zhang's Linkedin Analytics