Jarrett Evans Email and Phone Number

Work with Business partners to understand components of their current data processes to find opportunities to streamline their data flow to help efficiently add value to their team.• Produced a P.O.C. of a specialized LLM chatbot that leverages a RAG based architecture. This chatbot is capable of semantic search, historical summary generation and novel content creation.• Developed a Python pipeline that leverages regular expressions, datetime functions and calculations to do transformations on survey metadata. This data is then published in Tableau reports for executive leadership to help predict various market movements and to assist with policy decisions.• Produced a workflow in Python and Tableau that visualizes the liquidity of numerous financial assets around the world. Worked with many stakeholders to understand the process of how these assets are measured currently and how we can take steps to help streamline the process to make it faster and more flexible.• Created web scrapers in Python using the Beautiful Soup library to extract and parse speech data from various Federal Reserve district bank websites. The data generated was preprocessed to be input for machine learning NLP algorithms and reporting metrics developed by the team’s data scientists.• Developed a strategy for the storage of geospatial climate data utilizing AWS S3 buckets and Starburst to fall in line with a data mesh infrastructure.• Implemented a PCA analysis on markets-based liquidity data to assist with future monetary policy decisions.• Converted a Matlab code base to Python to assist with automating the process of options reporting. These python scripts are fed into a dash app to create a faster time to analysis.

Used this career break to gain more domain knowledge in Oncology. The primary purpose was to better understand how data science can be applied to common cancer research and cancer care issues. The channels utilized for studying consisted of textbooks, research papers, podcasts, and lectures. The following topics were studied to get a baseline level of comprehension and to be able to more effectively communicate with medical professionals.Bioinformatics• Clustering algorithms• Combinatorial algorithms and pattern matching• Computational proteomics Statistics of Clinical Oncology Trials• Statistical design, conduct, and analysis of clinical oncology trials• Development of endpoints• Monitoring safety and efficacy• Adaptive designsMolecular Biology of Cancer• Regulation of gene expression• Tumor Suppressor genes• Cancer Stem Cells• Tumor ImmunologyInterpretable Machine Learning• Model-agnostics methods• Image classification feature importance• Mitigating the bias of models and datasetsMachine Learning in Breast Cancer Research Blog Post• Evaluated a research article ‘Morphological and molecular breast cancer profiling through explainable machine learning’ to learn about the complexities of cancer research and how machine learning can act as a driver for the research.• Developed a blog post that highlights how machine learning can be used in oncology research (breast cancer specifically) for the purposes of more refined tumor grading and new candidate lists for potential targeted therapies.• Displayed an ability to understand the complexities of two different domains (oncology and machine learning) and communicate the results of the study in a digestible manner. Link to article: https://towardsdatascience.com/machine-learning-in-breast-cancer-research-d1dd99f7819a

Collaborated with stakeholders, who varied from project managers to engineering leads, from teams inside the organization to understand their core data needs and implement solutions to help them derive insights from their team’s data. (Contracted via Crystal Equation)• Built out multiple dashboards from the realms of team execution, user behavior, and product performance. Utilized SQL to create the queries that populated the visualizations with a Unidash UI. These dashboards allowed a user to see high-level metrics for reporting and provided the option for lower-level views to allow for precise decision-making.• Assisted with the planning of machine learning pipeline design in the team’s predictive analytics branch.• Used programming notebooks to explore text-based error data through machine learning methods such as NLP, classification, and clustering. Leveraged the python data science libraries Pandas, NumPy, Matplotlib, and scikit-learn.• Helped begin an initiative to provide standardized product analytics for products the client teams produce. The standardization allowed the user to compare how various products perform more easily.• Monitored and built ETL pipelines with a Python framework, Dataswarm, to fill a team-specific data repository. The pipelines populated a data mart with aggregated versions of reporting metrics for faster dashboard development.• Mentored a team of 12 associates across 3 continents to help them develop their skills within the data analytics space. These skills included SQL query writing in Hive, dashboard building with Unidash, and statistical interpretation.• Provided various analytics to help with marketing initiatives for the client team to help enhance their product adoption within the company. Some analytics provided included churn rate, power users, and time to perform a task. A newly launched product saw exponential growth in number of users (about 300%).

Delivered key information from different types of data (e.g. transactional, supply chain, time series) to numerous business managers. Communicated with business partners to understand logic and domain knowledge of different processes so that the reporting team could improve reports and provide valuable recommendations. • Created and monitored a collection of 12 ETL pipelines in Alteryx. These pipelines allowed for easy access to data commonly requested for AD-Hoc reports and automated various reports that would be provided daily to business partners. • Automated a manual report generation process with multiple sequential-based aggregation functions in Alteryx. The automation saved the business partner at least 8 hours each month.• Designed an interactive dashboard that gave business partners access to contact channel distribution data through Alteryx and Tableau. This provided managers the ability to determine how customers tend to initiate different services the company provided.• Built a model that forecasts the amount of volume for every operations process each month using Arima and ETS forecasting techniques. This aided the manager's preparation efforts for staff allocation on a month-to-month basis. • Conducted a workshop series on Hadoop data marts, covering topics such as table creation, configuring data connections, the functions of different data types, and how different joins work. After the 3 workshops, the 7 attendees had a better understanding of how data migration occurred across the team.

Assisted the service department by creating analytical insights into the different pricing aspects of 3D printers.• Utilized data collected from multiple departments to create a math driven outlook on the full cost of selling a printer. Variables included scheduled maintenance, the likelihood of parts breaking at given times, and field engineer billing costs. This approach allowed for more appropriate printer prices along with the length of warranty for the machines.

Helped to ensure that the independent financial advisors were performing within the ideals of LPL Financial.• Reviewed types of communication from independent advisors such as emails and advertisements. The validation that these communications were within the company guidelines helped the company avoid potential lawsuits.• Verified commissions that would be submitted to check that the advisor had the proper licensing. This helped to continuously build trust between LPL Financial and the clients of the independent financial advisors.