Shelly Yaffe

Professional Experience

Cat Digital

Cat Digital is the technology entity of Caterpillar construction company.

Data Engineer - April 2023 - present

Technologies Used: Python, Snowflake SQL, AWS: Lambda, SNS, SQS, S3, Event Bridge, SageMaker, Kinesis, RDS, CloudWatch; Github, Github Copilot; AzureDevOps CICD

Developed data pipelines using python and AWS components to apply transformation logic and deliver customer assets to front end teams
Maintained and optimized existing ETL pipelines, identifying areas for improvement or retirement, while ensuring seamless delivery of client-facing data to the fleet management system
Wrote Sagemaker scripts to ETL data in preparation of a one time only insertion after new pipeline development was deployed to production to handle historical data backfill
Wrote API script in python to upload dealer asset records in bulk to save support team time from manually inserting data
Wrote regression, unit and integration tests based on behavior driven development best practices
Collaborated with technical owners to understand solution designs and translate the work into requirements such as mapping documents and test scenarios
Pioneered scrum and agile best practices for a team previously working in waterfall
Created technical design diagrams to illustrate expected flow of data through AWS components
Demo development milestones to engineering department

TISTA Science and Technology

TISTA is a government contractor focusing on data technologies.

Data Engineer - January 2019 - April 2023

Technologies Used: Python, Spark / Scala, AWS: EMR, S3, EC2, ECS; Github, Zeppelin, Camunda, Informatica: Power Center, Test Data Manager (TDM); Microsoft SQL Server, Docker, Airflow

Create Informatica mapping and workflow to pre-process, clean and validate legacy financial data
Develop proof of concept to mask personally identifiable information using Informatica Test Data Manager (TDM) for lower level environments
Develop Informatica solution to flag items reconciled but not reported (IBNR) against general ledger data
Build data warehouse of qualitative and quantitative data for actuary analysis
Monitor logs for monthly IBNR process and handle failed records appropriately
Demonstrate functionality of Informatica TDM tool and proof of concept to stakeholders
Develop Informatica mapping to reconcile financial data from various sources to a general ledger
Collaborate with subject matter experts across the enterprise to understand disparate data sources
Analyze homogenous data to learn relationships from different sources
Develop Spark / Scala programs to pre-process new cost measure data for Quality Payment Program
Assess and update current cost measure code to handle new requirements
Wrote unit tests to ensure expected data and edge cases would be processed accordingly
Wrote a python Airflow script to post and push final cost measure data to the QPP API
Support Data Scientists in troubleshooting abnormalities found in final datasets
Participate in code review sessions with Data Scientists to provide clarity on code changes based on requirements
Configure cost measure scripts to run on EMR clusters efficiently based on a Directed Acyclic Graph
Design and document case management architecture to support a scalable engine using AWS
Develop proof of concept for back-end BPMN engine to support case management process
Develop proof of concept to demonstrate Camunda authentication and authorization through REST Api calls
Design a business process model based for the Medical Review Post-pay process on collaboration stakeholders

CivicScape

CivicScape is a startup with one goal: put police officers in the right place at the right time. We bring the power of machine learning to the intersection of identifying where police presence would impact the decrease in criminal events.

Data Engineer - July 2017 - November 2018

Technologies Used: Python: Pandas, Numpy, Scikit-learn; R, PostgreSQL, AWS: EC2, S3, RDS, CloudWatch; Github, Linux

Evaluate model selection to generate risk scores for where and when crime is most likely going to happen
Develop solution to generate model performance and visualizations
Run monthly penetration tests to ensure application security was not compromised
Evaluate model improvements using disparate features and inputs
Develop tool to produce and archive historical risk score rankings in AWS RDS
Optimized data loading into RDS tables from days to seconds utilizing python 3 and asynchronous methods with asyncio
Preprocess and transform crime incident reports and 911 calls for service into a temporal and spatial dataset
Implement solution to routinely ingest and surface disparate data sources from clients i.e. treatment effects and security cameras
Initiated unit tests suite across all modules
Convert US census tract data into two dimensional matrices to be used by neural network for spatial risk scores
Create databases for research projects and run validation checks to ensure clean data
Research and solution proof of concepts for potential portfolios
Maintain multiple RDS databases that house 10+ million rows of data and monitor cloud watch error logs
Transformed variety of spatial & temporal data in order to provide the CrimeScape neural network with an optimal signal for best performance (i.e. business licenses, holiday and events schedule, treatment data, etc.)
Provide weekly client support and communication to active users

GCM Grosvenor

GCM Grosvenor is an alternative investments firm. In the 8 years I was there, it evolved from a fund of funds into an alternative investments firm with a focus on giving clients transparency to their portfolios.

Software Engineer - June 2015 - July 2017

Technologies Used: C#, ASP.NET MVC 4, KnockoutJS, JavaScript, JQuery, HTML, CSS, SQL, MS Azure (Table Storage & Blob Containers), NServiceBus, NUnit, Github, TFS, NUnit, RhinoMocks, AWS Lambda, Tableau

Develop and support multiple applications that power suite the Client Facing Systems which resided in Microsoft Azure using blobs and table storage
Develop new features and maintained multiple ASP.NET MVC applications using Azure NServiceBus to send, receive and process messages
Run frequent load tests to ensure stable application performance
Implement features on Client Facing applications to surface qualitative and quantitative data giving clients transparency to portfolio return and performance
Setup and configure CI(D) environment using GOCD and TFS
Write code with Test Driven Development in mind using unit and acceptance tests to ensure production-ready code
Developed a seven step guided tour of all features on a newly introduced return analytics application leveraging the open source library intro.js
Worked closely with the UX team to ensure I was coding a product with the user in mind
Provided support for reconciliation efforts to validate expected returns and balances
Supported ETL packages for IT budget data source and troubleshoot cross-system bugs
Create SQL audit agent program to receive and send available audit files to Amazon S3 client

Executive Assistant to the CTO

Provided support to the CTO 24/7 for operational and logistic requests
Created and implemented IT Budget control process using Nintex workflows
Captured metrics to show reduced expenses in IT spend overtime

Assistant - Software Development

Prepared business work flows for department and firm onboarding process to enable automation
Designed and created SSRS reports to show department wide resource allocations
Created and supported demand planning reports for product owners to track respective department work requests

HR Generalist

Wrote business requirements for PeopleSoft Self Service launch & provided tier one technical support to users
On boarded all new hire employees and contractors
Provided technical support for the performance management system

shelly.yaffe@gmail.com