David Maxson Email & Phone Number

Auburn, Ca, Us

New York, New York, Us

Business-to-business consultant delivering machine-friendly data feeds from human-oriented sources. Lead engineer designing and building most of the internal platform, approach, and developer tooling that our team and junior engineers used to deliver this revenue-generating service.- Built and maintained 50+ bespoke data pipelines, including delivering most of our team’s hardest contracts- Accelerated whole team’s pipeline authoring by 10x by building custom platform and approach, including building and publishing no-code data pipeline authoring and graph-defined container definition tools- Improved customer experience and reduced internal overhead and time to detect failures from days to minutes- Saved 10s of man-hours per week by streamlining developer workflow with custom CLI and automating system deployment using CI/CD and IaC- Enabled one-line solutions in non-automatable data pipelines by implementing AI/Human hybrid functions- Trained all levels of engineers on Large Language Model (LLM) usage and delivering Retrieval Augmented Generation (RAG) solutions to customers- Managed same-day customer communications about delivery issues, triaging issues, managing expectations, clarifying requirements, and implementing urgent fixes- Supported hiring initiatives, including performing dozens of senior-level technical and cultural interviews and offline evaluations

Laurel, Maryland, Us

Delivered near-real time analytics for >$100M Government (IARPA) project evaluating competing biometric security algorithms running over a novel dataset (~30TB). Received multiple awards for my critical contributions to the massive success of this project both within APL and at IARPA. Also served as a researcher and research engineer on a wide variety of other AI and data projects.- Reduced analytics delivery time from 1-3 days to near-real time by building custom Django and Vue.js web app, including building C# and Rust microservices connected with GRPCStandardized data management and precise analytics across multiple partner organizations in government, industry, and academia- Enabled reproducible and semi-automated data cleanup on the project’s custom dataset, catching hundreds of erroneous annotations, improving perceived performance of partner algorithms, reducing errors in analytical results, and enabling the project sponsor to make more confident decisions- Accelerated iteration cycles from 2-5 days down to minutes for multiple internal teams, while reducing burden on the analytics team by pushing interactive self-service analytics to end users- Quantified algorithmic efficiency of partner companies’ custom-built algorithms by training in-house deep convolutional neural networks (CNN) using Pytorch- Accelerated research iteration cycle from days to hours on a reinforcement learning-driven CNN project (in Pytorch) by building a data-defined experimentation framework- Recommended COVID resource allocation by implementing, scaling (Kubernetes), analyzing, improving, and presenting epidemiology models- Made code production-ready months sooner and expanded development from a few individuals to the entire analytics team by implementing a translation library that transparently migrated and scaled analytical code, without change, from the local dev sandbox into the customer’s highly-restricted production environment

- Prototyped novel decision engine that converted complicated, multi-person decision process over days or months into a real-time, quantifiable, explainable algorithm- Rapidly prototyped document retrieval engine for fuzzy, adaptive querying (semantic search)- Contributed to compiler for, and built a fully functional debugger on a system emulator of, a custom HPC ASIC processor to accelerate tool development and boost end-user productivity

Rotational summer intern with the Stokes internship program. Toured in computer science research (emphasis on neural networks), network security, and high performance computing.

Master'S Degree, Computer Science (Machine Learning)

Bachelor Of Computer Science, Artificial Intelligence

Computer Science