Key Skills

Software

Internal/Proprietary Tooling

Top Subject Matter

No subject matter listed

Top Data Types

Image

Text

Top Task Types

Classification

Data Collection

Evaluation Rating

Freelancer Overview

I contributed as a Data Annotator to an AI project that aimed to develop a unified machine learning model capable of assessing the data quality of customer by analysing the contextual text and images present within each data customer record. The main is to reduce both fraud risk and reduce high logistics cost due to inaccurate/incomplete physical delivery addresses. My tasks were as follows: Evaluating customers address data pulled from sales databases, based on metadata quality rules for example: Adress_line_n must contain unit/individual information such as company name, City must contain a valid city… Labelling address images provided by the customers and/or the Data scientist team in order to identify and classified address images. I use a UI tool developed internally to identify geographic location, breakdown multiple images contents in the same address , create key points to stake out the borders of the customer address, assign markups and labels to the entire image. In addition to these main tasks, I was responsible for customer data records noise cleaning id est records that contain information that is not classified as Address and labelling customers data that can be reused during last mile packages delivery for example POC phone number. In practice, a batch of customer data records and images were assigned to me and I was responsible for processing the batches on a proprietary UI dedicated to the project and submitting them within a deadline.

Labeling Experience

Education

Work History