Nicholas Seidl

● Computer Vision

● Natural Language Processing

● Data Science and Engineering

● Multi-modal (image, video, text) foundation models

● Data Curation - select the "best subset" from a larger dataset

● Cloud Compute

● Robustness Analysis

● Email me for most up-to-date resume

● Implemented YOLO(v3) neural network for live on-device object detection

● Optimized with quantization, CUDA, cuDNN, TensorRT, and Tensor Cores

● Achieved 80x speedup in inference time and up to 4x smaller model sizes

● Analyzed and visualized failures of neural face detector to target specific aggressors and guide development

● Curated datasets for different uses with custom model in the loop processes

● Managed data collection, selection, and annotation (guidelines, spec, and quality)

● Designed and automated failure analysis pipeline: inference → detection evaluation → failure analysis → visualization

● Content Collection: audio/image/video/stream analysis and retrieval

● Rebuilt web service via backend architecture and frontend upgrades

● Integrated blob (Amazon S3) and metadata (Postgres) stores to serve content

● Leveraged Python and Tornado for asynchronous and non-blocking REST API

● Extracted attributes from audio, image, video, and stream (HLS) content (Swift)

● Improved frontend with React, Bootstrap, Node, and Buildpack

● Image Comparisons: image comparison web service with history

● Rewrote backend in Python and Flask; expanded API for frontend and hosted comparisons

● Created minimal frontend with Typescript, React, Redux, Node, and Yarn

● Modularized backend so different comparison algorithms can easily be added

● Engineered pluggable Blockchain platform for transaction reconciliation and facilitation (big data)

● Designed, implemented, optimized baked and middle-tier pipeline (Python, Bash): MySQL → .csv → JSON → Elasticsearch → Blockchain → Hash Enforcer

● Leveraged Docker, Openshift, and Redhat for rapid deployment of system infrastructure

● Scrum Master in Agile environment for a team of 8 developers (1 week sprints and iterations)

● Created Shark Tank presentations bi-weekly: demo, business case analysis

● Presented to various high level employees including: Directors, Presidents, CIO, CTO, COO

● Aid students and faculty in troubleshooting of various technology problems (faulty hardware, buggy printer and proprietary software)

● Fulfill general helpdesk tasks: Windows/Linux/Mac OS problems

● Developed on Agile team of 10: optimized cusotm barebones Linux file system for mass deployment on small embedded systems (LiDARs) to work properly with Quanergy Software (C++)

● Created and tested scripts to streamline testing and preliminary deployment procedures (Bash, Python)

● Improved raw data logging feature for LiDARS using Quanergy software; implemented naming and data logging contentions (XML, JSON)

● Networked IP cameras, LiDARs, INS together to create robust test infrastructure

● Worked on continuous integration for software team (Jenkins, AWS, Stash, git)

● Configured, tested, optimized LDAP integration with Atlassian and Google Software Suite, Zabbix, Dell Synology (Confluence, Jira)

● Installed Zabbix monitoring tool onto all company machines (Linux/Mac/Windows) along with SNMP reporting

● Troubleshooted proprietary software through the command line on Linux, Mac, Windows OS

● Researched for and constructed various top of the line computer systems specialized for different uses

● Stress tested, benchmarked, overclocked systems to maximize performace and reliability