Nicholas Bovee

Profile

ML Engineer experienced in full-stack AI; including vision, edge, & robotics domain knowledge. Proven track record of creating impactful tools. Seeking opportunities to build impactful ML&AI in unique engineering environments.

Experience

ML & AI Engineer
Rowan University, CS Vehicles Grant [contract]
Glassboro, NJ
February 2022 - Current
  • Led research cycles with 10 member team to rapidly prototype and deploy AI solutions, mentoring on optimization of artificial intelligence and machine learning models.
  • Architected production MLOps platform with zero-deploy framework supporting efficient allocation and monitoring of AI workloads on resource-constrained vehicular and edge platforms.
  • Achieved 7% inference speed improvement through performance profiling and optimization on embedded hardware, reducing energy consumption for sustainable edge deployments.
  • Technical Lead for team at NSF regional I-Corps, deploying ML models for boutique crop life-cycle applications.
  • Design and integrate bi-manual SO-100 arms into custom robot frame for hierarchical Vision Language Models.
  • Containerized ROS2 in Docker to maintain consistent dependencies across edge-devices in robotic vehicles.
Senior Staff Engineer
Rowan University, College of Engineering
Glassboro, NJ
November 2021 - Current
  • Architect multidisciplinary engineering solutions integrating full-stack development, embedded systems, and DevOps practices for Engineering ERP services using Django, Docker, and Dragonfly, serving entire 2,000+ Engineering college.
  • Lead development team in applying DevOps practices by integrating CI/CD checks for quality assurance in code submissions and revising system documentation across distributed micro services.
  • Create CI/CD tooling pipelines for front-end and back-end testing using Gitlab CI and GitHub Actions, increasing pre-production bug detection by 38% and reducing post-deployment issues by 21% annually.
  • Collaborate closely with teams from other disciplines to develop a flexible solutions for all stakeholders.
  • Create compliant access management with CAS SSO and student ID cards to validate student training records when accessing industrial equipment.
  • Feature rollout of 5,000-SKU inventory system with vendor API integration, accelerating throughput by 66% and automating all inventory requests measured over 24 months.
  • Develop automated part request management with integrated inventory tracking, tripling lab task throughput on over 2,500 high-touch tasks annually.
Staff Engineer
Rowan University, College of Engineering
Glassboro, NJ
February 2018 - February 2022
  • Deployed recurring calibration schedule for $500k+ of instrumentation, providing advanced lab capabilities.
  • Rapid development of custom embedded devices to support individualized research needs.
  • Managed 12 student employees to support simultaneous course operations across the College of Engineering.
  • Optimize engineering equipment utilization across 40 sections of engineering education with only two engineers.
  • Delivered annual training sessions to 120+ students on safe operating procedure for industrial equipment.
Automation Engineer
Ernie Green Industries
Rochester, NY
December 2015 - August 2017
  • Developed real-time image based defect identification that seamlessly integrated with existing processes.
  • Engineer robotic end-effectors to decrease injection molding cycle times by an average of 17%.
  • Delivered critical onsite operational integration for PLC mold interlock logic installations at international locations.
  • Implemented inspection systems, achieving $171k annual cost savings with defect reduction; ROI under <4 months.

Other Works

Course Scheduling RAG Agent [Internal Mockup]
April 2025
  • Rapid prototype demonstrating AI-first development of production-ready agentic RAG system.
  • Allowed for conversational control over student services, without direct user entry to underlying service.
  • Built custom prompt engineering pipeline with hybrid vector database to support course discovery.
  • Agentic workflow with secure API tools, multi-stage reasoning, validation & constraint satisfaction.
  • Full observability with tracing for production debugging, error analysis, and fine-tuning pipeline.
  • Input/output validation and monitoring to protect from malicious usage, hallucination, and identify user distress.
  • Schedule validation tool calls enabled minimal context windows, reducing API costs while maintaining accuracy.
Food Bank Management System [Hunger Free Campus Grant]
March 2024 - May 2024
  • Developed a robust offline network for pantry operations, utilizing Django & SQLite, serving 6000 students.
  • Integrated compatibility with UPC-A, UPC-E, and QR internal barcodes for digital asset tracking.
  • Introduced intuitive user interfaces to help new staff navigate inventory verification of 1000+ SKUs seamlessly.
  • Configured embedded Linux builds with Pi-Gen, ensuring system compliance and efficient resource usage.
Reinforcement Learning of Street Fighter Training Partners [MS - Machine Learning II]
January 2023 - May 2023
  • Evaluate capability of Reinforcement Learning trained AI opponents gated by human physiological limitations.
  • Agent constrained to naive basic inputs, to organically discovering advanced techniques through exploration.
  • Reward function considered indirect advantages to strengthen agent’s approach. Severely penalty on loss.
  • Asynchronous Advantage Actor-Critic used to optimize resource allocation during training.
Zero Shot DNN Behavioral Evaluation [MS - Machine Learning I]
September 2022 - December 2022
  • Constructed and analyzed Zero Shot CNN classifier on the Caltech-UCSD Birds-200-2011 dataset in PyTorch.
  • Initially trained on attribute segmentation data, enabling feature-specific classification of known species.
  • Fine-tuned additional layers to identify species held outside training set.

Skills

Preferred Programming LanguagesPythonGo/GolangRustTypeScript
ML & AIPyTorchLangChainDSPyOpenAI APIAnthropic APIscikit-learnMLFlowSGLangvLLM
Data & Vector EmbeddingsPandasChromaDBpgvectorElasticSearchPostgreSQLSQLite
WebDjangoFastAPIgRPCRedisDragonflyDramatiq
DevOps & ToolingDockerKubernetesNixAWSGithub & Gitlab CICD

Education

Computer Science
Master of Science
3.94 GPA
Rowan University
2023
Upsilon Pi Epsilon 2022 & Outstanding Graduate Award 2023
Mechanical Engineering
Bachelor of Science
Rochester Institute of Technology
2015
Co-ops at General Dynamics & iRobot Corporation

Publications

SplitTracr: A Flexible Performance Evaluation Tool for Cooperative Inference and Split Computing
ICPE
2025
N Bovee, I Ali, G Patapanchala, S Bitla, SS Ho
SplitTracer: A Cooperative Inference Evaluation Toolkit for Computation Offloading on the Edge
ICFEC
2024
N Bovee, S Bitla, GK Patapanchala, SS Ho
Computation Offloading for Precision Agriculture using Cooperative Inference
ICFEC
2024
S Bitla, K Patapanchala, N Bovee, PR Sanchez, SS Ho
Experimental Test-bed for Computation Offloading for Cooperative Inference on Edge Devices
SEC
2023
N Bovee, S Piccolo, SS Ho, N Wang