Maria Seirli

As Data Science Manager in Accenture, I have been involved in various projects and industries including Retail, Resources, Mining.

I have strong functional expertise in Data Analysis & Curation, Predictive Analytics,Customer Segmentation, Text Mining, Process Optimization,Quality Prediction acting as a Tech Lead and Project Manager.

I am specializing in end-to-end automated solutions using advanced AI methods and big data analytics for various use cases supporting clients worldwide for the last 5+ years.

•Acting as Tech Lead the development and deployment of the solution in Production for the analysis of production optimalization & operations planning through predicting operation’s risks in mining using IoT data

•Lead the curation and the transformation of the data using ADF & Databricks

•Employed Advanced ML techniques in Azure ML to predict the unexpected events of the process based on Geological and Operational conditions.

•Lead the developed of recommendations engine targeting on safety & value maximization allowing the user to interact with various what if scenarios.

•Lead the design of the customized application with automated processes built in Azure for Mining Production Reporting, Simulation Scenarios, Strata Risk Predictions assessment and Mining Production Optimization.

•Regular Stakeholder Management and reporting during governance meetings

•Acted as Technical Lead for developing a Big Data capability for the customer that would create alerts for the Fault Detection & Classification for an equipment using Spark Streaming ,for analyzing batches of data using technologies of Azure cloud for ingestion and curation and Databricks.

•Guided the development of the KPIS in cloud used for identifying issues and created alerts using Azure Event Hub.

•Coordinated the team (tasks allocation, monitored velocity & releases). Managed issues & mitigated data challenges by proposing data optimization approaches in the code.

• •Leading statistical modeling analysis for a global search engine company. Delivered an end-to-end automated solution for predicting churn. The solution has been developed in Python and R and helps client to identify and target high value customers with probability to churn at an early stage of onboarding. Propensity modeling has been applied on a segmentation scheme created for company’s acquisitions. Various ML algorithms were tested such as RF, SVM, Logistic to select best fitted model with regards to business requirements. The solution has been integrated into company’s systems in order to be ready to run on a production mode.

• Ensuring data quality for Upper Trade and applied checks for data integrity.
• Leading the estimation of the Greek market for lower trade
• Applied sampling and forecasting of economic method for estimating trend for retail stores in greece