Software Engineer

Marcus Agard

Software Engineer

New York, NY -- 347-***-**** -- **********@*****.***

https://github.com/maragard -- https://linkedin.com/in/marcus-agard

Technical Skills

Programming Languages -- Python, Node.js, SQL, Typescript, Bash, Powershell, Rust, Golang, Java, Scala

Networking -- Wireshark/TCPDump, DNS servers, socket analysis, Snort, virtual networking w/ VirtualBox

Frameworks -- Selenium, Sk-learn, Tensorflow, PyTorch, Flask, Django, Pandas, React, Docker, GraphQL, Next.js, Pytest, Jest, React, React Query, Vite, Playwright

Tools -- Jenkins, Ansible, AWX, Kubernetes, git, Terraform, CDK, Sonarqube, Databricks, Red Hat OpenShift, Claude Code, Github Copilot

Operating systems -- Windows Server 2019, 2022, 2025; Linux distributions Debian, Ubuntu, Kali, & CentOS

Work Experience

PNC Financial Services Group, Software Developer II October 2022 - December 2025

●Authored robust automation workflows leveraging Python, PowerShell, and Ansible, streamlining software runtime by 80%.

●Spearheaded the development and delivery of 5+ fully functional automation modules, exceeding individual project goals by 20% and boosting overall bank productivity by 90%.

●Eliminated 10+ vulnerabilities in team's API and React front-end by implementing strict input validation and output encoding, achieving a 99.99% security score.

●Mentored and supervised 2 intern teams in developing full stack projects for team leadership, resulting in the onboarding of two interns to part-time roles and 200 hours of time savings.

●Engineered ranking model to assist users with platform, achieving a 30% reduction in platform onboarding time and an AUC of .9.

●Championed a data-driven approach to sprint planning, analyzing past performance metrics to identify bottlenecks and optimize resource allocation, accelerating feature delivery by two sprints per quarter.

Huntington Bank, Machine Learning Operations Engineer July 2022 - September 2022

●Collaborated with vendor AWS developers to engineer new infrastructure for ML model deployment, leading to 75% faster model deployment for 200+ business teams.

●Composed Terraform modules to provision and manage AWS resources through Azure DevOps pipelines, improving infrastructure deployment speed by 40% and reducing deployment failures.

●Shifted team processing to AWS Lambda, reducing cost overhead by $1000s.

●Spearheaded the development of an NLP model to automate the extraction of key information from financial documents, decreasing processing time for loan applications by 40%.

FINRA, Machine Learning Operations Engineer April 2021 - May 2022

●Expedited development support for product teams with an average ticket closure in 2 days.

●Architected solutions addressing user needs across all FINRA technology teams within a heavily restricted cloud environment.

●Piloted a Kubernetes Proof of Concept (PoC), assessing the feasibility of integrating new technologies for container orchestration and reduced deployment times by 40% in the test environment.

●Enhanced existing Python and Java applications, boosting computational efficiency by 60% and reducing cost by hundreds a month.

Rubbish, Data Specialist September 2017 - August 2020

●Optimized both SQL and NoSQL databases to improve app responsiveness by 90%.

●Reduced data visualization build times by 85% through web API.

●Automated deployment processes through CI/CD pipelines across AWS and GCP, accelerating software release cycles by 60% and reducing deployment errors by 35%, while improving developer productivity.

●Achieved a code coverage of 90% for new Apple Vision machine learning features by implementing robust unit and integration testing frameworks and continuous integration pipelines.

●Deployed models with mAP of .7, and precision of 89%

Technical Projects

SBSCT (Pronounced Seabiscuit)

●Engineered a neural network for predicting horse racing outcomes using Python & scikit-learn, achieving 86% accuracy and 90% precision.

●Owned all model development from data cleaning to deployment.

●Developed a data pipeline to ingest live horse racing data at 40GB/day, enabling real-time predictions.

●Processed 400k data points of historical horse racing data for feature engineering and model training.

●Implemented PCA for dimensionality reduction, optimizing model performance with a 66% reduction in features.

Contact this candidate