Vannel Zeufack Email & Phone Number

Atlanta, Ga, Us

Alpharetta, Georgia, United States

• Leads a team of 3 to build and maintain quality data pipelines.• Work closely with client facing team to gather requirements, set project timelines and support all clients (10+) data ingestion needs.• Collaborate with product team for sprint planning and project tracking on Zoho.• Collaborate with data science team to provide actionable customer data to support analytics efforts.• Conducted 40+ technical interviews which allowed to fill 5+ roles in the ETL and data science teams.• Championed thorough code reviews within the team, resulting in improved codebase quality.• Orchestrated the seamless transition of all our customers to an enhanced data model, enabling more robust analytics.

Alpharetta, Georgia, United States

• Designed, implemented, and maintained 30+ ETL pipelines across all customers using ECL (SQL like language), HPCC Systems, Kafka and ELK, allowing a smooth intake of millions of transactions daily.• Increased data quality by 15% through data validation and cleaning.• Reduced the overall ingestion time by 30% by identifying and optimizing resource consuming jobs through optimal use of indices.• Enhanced code quality by implementing clean code standards and organization leading to a 40% decrease in total lines of code required for an ETL, fewer bugs and faster code reviews.• Created training documentation used to train all new ETL developers.• Implemented data masking on personal identifying data to ensure client’s data security.• Collaborated with IT Ops to support deployment needs and resolve production bugs.

United States

• Used ECL to develop a machine learning preprocessing bundle, aimed at facilitating the data preparation phase of machine learning projects developed on the HPCC Systems platform.• The bundle included 7 modules: label encoding, one hot encoding, standard scaling, min-max scaling, normalization, percentage-based data split and stratified data splitting.• The module adhered to clean code standards, including fixed line length, consistent indentation (2 spaces), appropriate comments, and proper naming conventions. It underwent successful functional and performance testing, handling over 50 million records.

• Developed a K-Means based anomaly detection system for detecting potential abnormal users and potential abnormal periods of activity in a web server by analyzing the logs.• For detecting abnormal users, 7 features including average daily requests, number of active days and average size of requests were extracted. The users were then clustered to flag abnormal ones.• For detecting abnormal periods of activity, the logs were divided into one-hour periods to determine the count of requests per period. Periods were then clustered, and abnormal ones were flagged.• 3rd place winner at the Summer 2019 HPCC Systems Poster Competition

Master In Computer Science, Computer Science, 4.0

Bachelor Of Science - Bs, Computer Science, 3.83

Activities and Societies: Peer mentor for International Peer mentor program and the Odyssey Program