About Me

My name is Daniel Marley, I'm 26, currently based in Philadelphia, and a full-time, full-stack Software Engineer at J.P. Morgan Chase. I've been working full time since graduating from The George Washington University in 2020 with a B.S. in Computer Science and recently completed my Master's degree in Artificial Intelligence at The Johns Hopkins University.

AI has been a strong interest of mine since my time in undergrad, where it was my area of focus in my major. Looking forward, with my Master's degree in hand, I am eagerly searching for projects where I can apply my skills. I want to leverage AI to solve real-world problems that have meaningful impact, particularly in areas of sustainability and health.

Outside of work, I stay active in local recreational leagues for volleyball, dodgeball, and kickball. I've recently returned to wheelthrowing as a hobby and also like to kayak on the Delaware River when I can. I find each of these activities not only deeply rewarding and but also a balancing to my work.

Experience

JPMorgan Chase — Software Engineer III

2023 - Present

Card Acquisition and Marketing Platform

  • Worked across two teams as a full-stack engineer, providing DevOps expertise as well as designing and implementing new features
  • Launched a new, in-house A/B testing framework to drive new marketing user experiences; generated an annualized increase in revenue of $60mm within first year, exceeding projections
  • Served as coordinator between business, design, analytics, and product teams in relation to releasing our initial set of tests and customer deliverables
  • Lead several experiments in design and implementation
  • Developed and presented a proof-of-concept AI system to manage A/B testing for company-wide hackathon

Lockheed Martin Space — Software Engineer II

2021 - 2023

Team 3

  • Joined legacy program to help guide modernization, including move to web displays from Windows Forms, containerization, .NET version migration, move to microservices from monolithic architecture, and shift to CI/CD
  • Addressed bugs from testing and operations as well as driving new development
  • Quickly became a lead on the team for MVC development, transitioning multiple displays and major functionalities to web
  • Lead spike to containerize supporting services as first step to dockerization

Lockheed Martin Space — Software Engineer I

2020 - 2021

Team 2

  • Joined a small team of 2 for short 4 month spurt to help cover immediate needs in full-stack development, focusing on front-end
  • Mature program with microservice architecture built for commanding users of a specific space asset
  • Assisted with live demos and tests for the customer and stakeholders

Team 1

  • R&D project being brought to production with the goal of simplifying satellite commanding for multiple different asset types with limited resources
  • Modern microservice architecture with Angular frontend deployed to Kubernetes cluster
  • Acted as a lead for UI on a small team of 4, led charge to revamp frontend for microservice while still building out core functionalities for the product in a full-stack capacity
  • Presented directly to the customer and shareholders on several occasions for demos and major milestone testing
  • Spearheaded networking and security for the application's deployment as well as completed the end-to-end test automation

Projects

Multi-Modal Model Research in Chest X-Ray Data

  • Extended research of an existing paper that built a multi-modal transformer-based model to accurately diagnose chest conditions based on scan and patient data
  • Explored boosting clinical accuracy by altering the architecture of the model, aiming to improve the consideration of patient medical reports in classification

FactCheckLLM

  • Developed and tested an LLM system based around a plugin to identify, research, fact check and ultimately annotate significant claims in a user's browser; based on the open source Llama3 model family
  • Tested against local benchmark framework, marking performance against Politifact's previous publications, the FEVER (Fact Extraction and Verification) LLM Benchmark, and the AVERITEC (Automated Verification of Textual Claims) LLM Benchmark

Hi-YAH (Help You Already Have)

  • Participated in hackathon to develop LLM based HR chatbot designed to help employees with disabilities explore resources available to them at their location
  • Built with use of OpenAI API and collection of HR resources we fed to the model based on relevance to user's initial prompt

CAPTCHA breaker

  • Fuzzy logic AI system to parse and decode text-based CAPTCHA security measures
  • From scratch methods of image processing and feature extraction alongside ML model that translates human heuristics and rules to a complex decision boundary

Artificial-Intelligence Bias Detection and Reduction

  • Generalized executable designed to detect racial bias within datasets and pre-process the data in a way to effectively reduces learned bias within a model without significant impacts to efficacy

VVOIP (Verified Voice Over Internet Protocol)

  • Drafted RFC and protocol to secure Voice Over Internet Protocol (VOIP), enforcing authentication to prevent number spoofing and reduce success rates of spam calls
  • Implemented working prototype phone and implementation through Angular and deployed onto AWS