Account Closures Analysis Dashboard
Built an interactive Power BI dashboard at 1st Source Bank to investigate rapid account closures and identify patterns across account opening trends, closure rates, onboarding, officer relationships, e-student accounts, returning account behavior, and transaction activity. Queried data using SQL, modeled it with DAX, and presented insights to technical and non-technical stakeholders to support retention and operational decision-making.
Power BI
SQL
DAX
Data Modeling
Business Intelligence
Stakeholder Communication
UD Chassis Project — Engineering Automation & Product Innovation
Contributed to Lippert's first-of-its-kind Upper Deck Chassis project for the RV manufacturing industry through R&D, LEAN engineering, prototype design, fabrication, installation, testing, and product launch support. Programmed a custom engineering calculator using Python, Excel, AutoCAD, and DraftSight to standardize modified rig calculations, connect engineering data to production materials, reduce manual errors, improve safety, and help achieve an approximately 34% production efficiency increase.
Python
Excel
AutoCAD
DraftSight
LEAN Engineering
R&D
Manufacturing
Radiation-Induced Fault Characterization in Sequential Logic
Investigated how radiation-induced charge disturbances affect the reliability of sequential CMOS logic by modeling transient current injections in a 4-bit master-slave D flip-flop shift register. Used LTspice simulations to analyze single-event transients, single-event upsets, node-level sensitivity, clock-dependent vulnerability, and master versus slave latch susceptibility.
LTspice
CMOS Logic
Sequential Circuits
Transient Injection
Fault Modeling
Microelectronics
Reliability Analysis
Gas Adsorption Studies of MCM-41 Porous Material
Conducted molecular physics research through the GROUPS Summer STEM Initiative studying helium confinement in MCM-41 mesoporous silica. Gained hands-on lab experience in material synthesis, academic literature analysis, experimental research, and technical writing while contributing to a broader IU Physics research effort on gas adsorption in low-dimensional systems.
Molecular Physics
Gas Adsorption
MCM-41
Research
Technical Writing
Lab Work
National AI Policy Framework for a Developing Global Economy
Designed a national AI policy framework for a fictional developing country, focusing on responsible AI adoption, agricultural modernization, ethics, governance, data rights, education, workforce development, and public-sector AI. Compared global AI policy models and delivered a policy brief, presentation, and press-style defense.
AI Governance
Policy Design
Risk Analysis
Ethics
Comparative Policy
Strategic Communication
Romi Autonomous Vehicle
Autonomous and manual control system for a Pololu Romi robot using closed-loop control, state machines, embedded C/C++, Python tooling, I2C communication, PID line following, encoder feedback, and data logging.
C/C++
Python
I2C
PID Control
Robotics
Embedded Systems
FPGA Digital Design Accelerators
FPGA hardware acceleration projects using SystemVerilog, AXI, DMA, and hardware/software co-design concepts. Focuses on digital design, acceleration, embedded systems, and performance improvements over software implementations.
SystemVerilog
FPGA
AXI
DMA
Digital Design
Hardware Acceleration
STM32 Embedded Systems Project
Embedded systems project using STM32 hardware, firmware, peripherals, system integration, and enclosure design. Includes hardware-aware design decisions such as standoffs, sensor access, moisture protection, and connector accessibility.
C
STM32
STM32CubeIDE
Firmware
Fusion 360
Embedded Design
ML Energy Load Microservice
Cloud-deployed machine learning microservice for predicting building energy loads using a Random Forest model and Flask API. Built around deployable inference, data analysis, and microservice architecture.
Python
Flask
Random Forest
Machine Learning
Cloud
Microservice
Embedded Neural Network Optimization
Optimized a Raspberry Pi-based neural network for human activity recognition, improving accuracy from approximately 0.80 to 0.95 while balancing inference speed, model size, and embedded performance constraints. Deployed the model using TensorFlow Lite for edge-device inference.
Python
TensorFlow Lite
Raspberry Pi
Machine Learning
Edge Computing
Bone Fracture Detection CNN
CNN-based medical imaging model for bone fracture detection. Applies computer vision, image preprocessing, deep learning, and healthcare-focused machine learning to support diagnostic image classification.
Python
TensorFlow
CNN
Computer Vision
Medical Imaging
Deep Learning