Valerie Lynch
Nominated Award: Women in AI Person of the Year
Linkedin profile of Person: https://www.linkedin.com/in/valerie-lynch-1303b74/
Valerie is responsible for leading ESB’s team of data scientists, defining central policies around data science and visualisation as well as leading ESB in the adoption of AI and ML technologies.
Having graduated from CIT in 2000 with a Bachelor of Science degree in Computer Applications, Valerie joined RCI a division of the Wyndham Worldwide Corporation in 2002 as a real time data analyst and quickly progressed through the company to a Team Lead role by continually showing initiative and integrity in each position held. Valerie joined ESB in 2006 as a member of the Logistics team in the National Contact Centre. One of the successful projects that she undertook during that time was the development of an SAP Business Warehouse, that was used to update reporting and introduce new technologies used within the Contact Centre to provide Real Time Analysis on incoming call volumes and fault logging.
In 2012 she took the opportunity to join the Business Intelligence Team as a Data Analyst in Electric Ireland where she managed the introduction and development of Advanced Data Analytics to the Credit Strategy Team by providing essential insights and trends on the customer base and provided insights to aid the successful role out of Residential Strategy campaigns in 2016. She was also involved in the roll out of the use of Data Visualisation Analytics with tools such as Tableau and SAP Business Objects
She furthered her education in 2015 by completing a Postgraduate Certificate in Statistics in Trinity College Dublin.
Valerie was instrumental in the introduction of a Hadoop Data Lake and driving the use of advanced data analytics to improve business insight and performance. More recently Valerie also led the way in the rollout of the Analytics Agile Framework and methodology across ESB Data and Analytics and her learnings have supported its wider adoption across the whole team.
Reason for Nomination
Valerie has been providing data-driven, action-oriented solutions to challenging business problems across ESB for over 15 years. She is a business-minded data scientist with a demonstrated ability to deliver valuable insights via data analytics and advanced data-driven methods, she is also relied on as a key advisor across ESB in driving customer growth and data understanding. She has worked on a variety of business led analytics project such as Credit Strategy Segmentation where she analysed how Electric Ireland might improve its debt collection strategy through a better customer segmentation approach. Not only did this improve debt collection, but it would also lead to an improved customer experience. Customer-level statistics were created from chronological data for each customer and new summary variables captured the differences in behaviour between customers.
The impact for the business was the completion of a focused debt management team with an aim to design debt collection activities based on each customer segment and a data driven implementation of collection activities rather than the previous method of collection based on historical business methods. The success of this project resulted in improved timeframe to collecting outstanding debt and overall debt issue, engagement with customers on a proactive basis, leading to a much-improved customer experience. Valerie was asked by the Electric Ireland residential team to carry out analysis to understand the driving factors behind Repeat contacts across the Electric Ireland customer base. Valerie gathered data from multiple sources across the business which included SAP, Impala & SQL tables.
The Model was developed in Cloudera Data Science Workbench using R and a profiling table created a new level of detail for each customer denoting a customer element for repeat contact. The profiling elements could also be overlaid on to the customer data to provide visibility of the repeat propensity for various customer segments for each trigger. Understanding the customer journey for those who repeatedly contact improved the services provided for all customers so that Electric Ireland can place the critical and necessary resources where needed. Valerie was instrumental in the introduction of a Hadoop Data Lake and driving the use of advanced data analytics to improve business insight and performance.
Valerie was involved from the initial project ideation all the way to development of the production lake environment. She led the identification of source systems required to populate the Hadoop Data Lake cluster and migrated customer-based models including Churn and Lifetime Value to the new Hadoop environment. For the first time a truly complete customer 360 was made available to the Electric Ireland business ensuring that the development of AI models was faster as a result of the reduced effort in data extraction. Our Data & Analytics team have already seen real success in implementing a number of these models across our organisation – from Churn Prediction & Root Cause Complaint Analysis in Customer Solutions.
Valerie has continually worked with ESB business stakeholders to understand and identify business problems, she understands how to formulate questions that define the business goals that the data science techniques can target and find the relevant data that helps answer the questions that define the objectives of any data science projects.
Valerie also continually aims to support and help the Data Science team grow into one that is empowered to: 1. Continuously execute in line with the broader ESB Strategy 2. Show case best practice and promote innovation and operational excellence 3. Maximise the potential of existing assets (data & projects) and technologies 4. Work to maintain ESB’s competitive advantage through continuous innovation and learning in areas such as Machine Learning, AI & Robotics Valerie promoted and developed the adoption of an Analytics Agile Methodology framework for project delivery, a consistent approach across all of ESB’s analytics teams, focusing engagement on customer needs; allowing for the creation of business requirements and providing a methodology that has the flexibility to adapt to changing stakeholder demands as awareness of the potential of analytics grows. The methodology allows user requirements to be easily distributed across the hub & spokes and further enable collaboration across the multiple Analytics teams through Data & Analytics.
Valerie was key in leading the ESB Data & Analytics team to win Analytics Team of the Year at the Analytics Institute Awards in 2022. Valerie has led the establishment of an Ethical AI strategy in ESB, as a foundational pillar of ESB’s ethical data values, it enables ESB to address the varied, complex, and recurrent challenges / risks associated with AI projects. This strategy focuses on leveraging AI to both address key risks that result from the handling of huge volumes of data, but also the ethical challenges of potential applications. As AI has become an integral part to products and services, Valerie defined an AI code of ethics to ensure that ESB continues to leverage technology and data in a manner that reflects their core values.
Valerie continues to promote an Ethical culture in ESB with the use of AI & Data acting as a first line of defence, underpinning our governance and design frameworks. As someone who has worked in STEM for 20 years, Valerie has supported and promoted other women in Data Analytics over this time. Ensuring that her team is as diverse as possible is something that she felt Data Science could help with. She completed an analysis of all the job ads in Data Science and Analytics by reviewing gender bias. In Data Science job ads, she discovered there was a 50/50 split in masculine/feminine wording, and in Data Analyst job ads the wording showed to be 75% masculine wording. In those cases where the job ad showed to be more neutral in language Valerie was more successful in hiring and promoting female candidates. She currently manages a team of 10 people in Data Science with 7 women, demonstrating her passion in ensuring not only the correct person is hired for the job but ensuring the development of women in STEM (average % females in Data Science in UK is 30%)
Additional Information:
Valerie has also introduced a Data Quality Assurance standard to all Data Science projects to ensure that data values which are inputted to models during validation, training or otherwise are of optimal quality. As part of this quality assurance step, she has also integrated three types of checks to be performed before any deployment of an AI model which include technical validity checks meant to ensure that the suggested approach can be successfully deployed and supported by data engineering. Research review, where a peer helps another data scientist validate their process and recommendations and Scope / KPI validity check, done with the product owner, to ensure the approach remains within the scope of the project and satisfies the project’s KPIs.
Valerie has been very passionate about promoting Data Science and Analytics within and outside ESB. She is currently running ESB’s 2nd Data Science Hackathon, which will identify Solar panels through image detection, this will support Networks’ planning of the solar integration to the Irish grid. The previous Data Science hackathon was organised by Valerie was to help the business improve churn propensity and introduced business insights to those in ESB not previously familiar with this area. She is also responsible for the annual ESB Data Visualisation Hackathon which is in its 4th year and promotes the use of analytics and storytelling through visualisations. Previous datasets included storm outages and recycling sustainable stats. Within the analytics team Valerie is responsible for AI Community Updates, organising monthly webinars promoting Analytics and AI methodologies, technologies and use cases. Valerie has started a Yammer community where she and her team provide tips, answer data questions, and share information on the most recent activities within Analytics.
Earlier this year, Valerie was asked to be involved as a mentor for teams taking part in X-Ignition, an event organised by Dogpatch Labs in Dublin City, it takes place on a yearly basis where teams work on different challenges looking at key opportunity areas for ESB. Valerie’s unique view and insights from her experience in all areas of ESB and Data Science, ensured she was a valuable addition to the mentoring team.
Valerie has also presented at CIGRE NGN Machine Learning Event, where her presentation gave the audience a greater awareness on how machine learning can be a useful tool in the utility industry and at Smart Grids conference presenting on the analysis, she completed on Smart Meters to determine new offers of smart products and services and how data analytics helped introduce new rate structures for residential energy consumers.