Aaron's Portfolio

 Omdena-Sonocare project. This project is a telemedicine vitals signal processing.  We had great difficulty in gaining access to datasets which did not have academic only licensing terms.  Therefore we decided upon project deliverables, the customer Sonocare in Nigeria is required to build a dataset. The project takes Blood Pressure, Pulse, Respiratory Rate, and SpO2, and symptom inputs, using a video processing system, which then it reposted to a front end for a proof of concept. The project almost being complete, I mandated strict accuracy controls on these systems.  The tech stacks is included on my resume. The largest concern is that the Nigerian Naira has decreased in value 70% since 2020 making deployment on U.S. based cloud systems cost prohibitive and research is being done into local native cloud hosting, or on premises computing systems with load balancing.   UPDATE: What was done is complete modifications in phase II for a Flutter App, with FastA...

 This project was very lengthy.  It involved taking a request form of a location and materials and other constraints, and feeding the request into a program, that queried a complex Neo4J schema that took awhile to build, and then feed it into DSPy, which then build a custom Chain of Thought system using Gemini 1.5, This returned matching using NLP and Neo4J, and then taking this matching and doing locational web search.  Once the program was finished, it would spit out a PDF and an email which was usable to the business. The design of the program, was because the office requests at WID spent weeks doing disaster resource planning and searching.  The system we built greatly decreases the time to disaster response once an application is approved.

This is literally about 3 hours of reading if it's decided to read through everything.  It's read only version but it's very interesting material (I was made fun of by other graduate students for taking the time to learn this material, but I find it invaluable practical information).   This may be updated later.   Undergraduate Biotechnology presentation of Algae Based Biofuels:  https://drive.google.com/file/d/13bBzZWl0NwTq2qB8FYQSIQkuqgAYC0K9/view?usp=sharing  Very Interesting Course right here:  Advanced Automotive Electricity! (View Only) https://drive.google.com/drive/folders/1kmTYyrr-8_qVhDvfpVgaljs1fdUuPjOr?usp=sharing Hybrid Systems Coursework and Documentation: (View Only) https://drive.google.com/file/d/1DT8D9xwNbzCJEcN_8yrTguPUh1PCW4H_/view?usp=sharing Manual Transmissions (View Only) Toyota: https://drive.google.com/drive/folders/1-3XhYZ_Uye7Kyxy2lkO4yUgAMB8t2ShG?usp=sharing    A/C systems Toyota (View Only): https://drive.google....

 This post is some highlights from my early college Nanotechnology training, in which just starting college, I was training to be a clean room laboratory technician.  Here is some highlights. Coprinus mushroom video recording of Fluorescence Microsocopy I know how to operate a fluorescence microscope: https://www.facebook.com/745426647/videos/10150661948061648/ Here is me directing an atomic force microscope: Here is what the readings look like: Scanning Electron Microscopy which I have been trained how to use correctly. TEM (Transmission Electron Microscopy), which is the highest magnification possible. Fluorescence Microscope Composite graphic I built: Open image in new tab to read: IZON particle sizer sample: https://drive.google.com/file/d/0B82LaO7baJ5NOTRkYmY5ODQtZmRmOC00Y2M0LTlmODEtYTQ3MzcwZDFjOGQz/view?usp=sharing&resourcekey=0-0DSjPrY7kLjaJZPi1O5sOw  

 I spent about a year working on the Sutter Buttes, of California.  Although it's not as in depth as the classroom, I did learn a lot of things about safety culture, and gas production. Typically a Gas well is dug to 8,500 ft.  The way the rig operates is they use Halliburton's directional drilling tools, which allow more holes to be driller per pad (4 holes), this reduces environmental impact. Halliburton's drilling tools involve gyroscopic radio metric systems, which monitors the depth, speed, and at the same time, there is a group of people called "Mud-Loggers" who interact with the Geology of the gas-well, also measuring bore pressure using these installed instruments. Well Casing (with a casing crew), is usually put the well Casing to 2000 ft.  This prevents any chemical contact coming into contact with the water table.  As the chemicals may be hazardous to environmental health. Here is some photos of me and rig: The rig: Here is a photo of me on the rig ri...

 T-Ten requirement for passing T-Ten certification was passing 2 ASE exams.  I think it's a good program because it's an introduction into working on all kinds of vehicles.  The Toyota brand has a reliable hybrid synergy drive, which is important to understand. I have passed quite a few ASE exams, if you aren't aware of ASE exams what they mean let me explain for you. The ASE A-6 exam, also known as the “Automobile and Light Truck Electrical/Electronic Systems” test, evaluates a technician’s knowledge of electrical and electronic systems in vehicles. The exam covers topics such as wiring diagrams, circuit analysis, battery and charging systems, starting systems, lighting, and accessory circuits. It also includes diagnosing and repairing electronic control units (ECUs), sensors, and related components. Understanding of basic electrical principles, circuit troubleshooting, and use of testing equipment is essential for passing this exam. The ASE A-8 exam, also known as the “...

              My first foray into automotive systems was restoring a 1972 Chevrolet Nova, rebuilding the Small Block Chevrolet Power Train, the completed project is here. Additional Details: Bored .30 over, balanced pistons and crankshaft, federal mogul sealed power engine kit, hooker headers, Edelbrock 650 CFM 4bbl carburetor, .256 rear end, with 3-speed hydromatic, with 3" exhaust with an H-Pipe.  Had a really lovely sound, and had more power than traction honestly.     

  We had some difficulties in how to accomplish this project, plus I had budget shortfall, creating a dataset may be costly. We settled on a synthetic dataset and plan on bolting-on additional parameters upon correctly building a dataset. The data pipeline contains ingestion, validation, transformation, training, evaluation, and a prompt Front end. The biggest hangup was experimental design of the dataset, as well as finding localization UV/IR spectroscopy lab availability. This project is seeking funding to create a bigger dataset, the hard programming is complete, thanks to Partha Pratim Kalita , and Modeling by Suraj Mishra .   https://dagshub.com/Omdena/NorthCarolina_CameroonChapter_AngusIssues Dagshub Repository  Final Presentation youtube:   https://youtu.be/W_5Qh5oevq8?si=OnIf_JreGJI0mkBa  Corporate  LLC Blog release: https://swift-tree.dev/?p=257

  3D-MICE 3D-MICE (Multiple Imputation by Chained Equations) : A statistical technique used to handle missing data by creating multiple imputed datasets. The "chained equations" approach iteratively fills in missing values by modeling each variable with missing data as a function of the others, cycling through the variables until the process stabilizes. This method is useful in datasets with complex interdependencies among variables. Web Scraping   Web Scraping : A technique used to automatically extract data from websites. Web scraping is used to gather large amounts of data from the web for analysis, typically using tools like BeautifulSoup, Scrapy, or Selenium in Python. NLP Recommenders   NLP Recommenders : Natural Language Processing (NLP) is used in building recommendation systems that understand and analyze human language to make personalized recommendations, such as suggesting books, movies, or articles based on user preferences. Naïve Bayes   Naïve Bayes : A...

Computer Vision Fire detection Capstone Thesis Advanced Database Project (Coldfusion, PostGRESQL) (Unfinished) Adv Software Engineering Project (Front and Back End working with AI Recommender System) First place Nationwide Hackaton 2023 Spring MSc Computer Science Summary I am a graduate in Computer Science with a strong foundation in both theoretical and applied aspects of the field. My academic journey has been enriched by a diverse set of advanced courses and hands-on projects, culminating in a thesis focused on computer vision. Key Areas of Expertise: Computer Vision: Completed a thesis in computer vision, demonstrating proficiency in image processing, pattern recognition, and deep learning techniques. Advanced Databases: In-depth knowledge of database systems, including design, optimization, and the use of advanced database management systems. Advanced Algorithms: Strong understanding of algorithmic design and analysis, with the ability to develop efficient and scalable...

Calculus Project several algorithms with writeup Single Electron Tunneling Transistors! (Write up) Quantum mechanics TI-92 Formulas Surface Enhanced Raman Spectroscopy Writeup Stress Calculations of a Beam Simplified Silver NanoWires (Chemical Engineering) NanoModeling and Applications computer simulations  Here's my summary of the subjects you studied during your first year of nanoengineering: 1. Numerical Analysis Overview : Numerical Analysis focuses on algorithms for solving mathematical problems numerically rather than symbolically. This includes techniques for approximating solutions to equations, integration, differentiation, and optimization. Key Topics : Error analysis and numerical stability. Methods for solving linear and nonlinear equations. Numerical differentiation and integration. Numerical solutions of differential equations. Interpolation and approximation of functions. 2. Differential Equations Overview : Differential Equations involve equations that r...

ATM Machine -- ATM Machine Hosted Snow Table -- Snow Table Hosted Caldwell Spa and Resort no GUI program Caldwell Spa and Resort Hosted Decision Statements Program no GUI Hosted SoftwareSales Discounts Program BankCharges for transactions Hosted Faculty Raises program hosted BMI (Body Mass Index Calculator) BMI Calculator hosted Password Verifier Restaurant Bill Calculator Bank Charges and Shipping Calculator Bank Charges 2 project CIAA (Colored Intercollegic Athletic Association) dataset sorts Basic interactable Car and Television Objects Roman Numerals and Time Calculator Sorting Algorithms Test Scores and Averages Dice Simulator and Stats Calculator Netbeans GUI Store Project Deliverable Phase II PhP Example CIAA Another Demonstration of mini dataset   --------------------------------------------   Data Structures Folder (Too big to Host)

Here is a link to a comprehensive document regarding the project Link to the document Here is a video demonstration of it's usage: Link to the video Here is the project, it became open source: Dagshub Repo

 In this project we completed computer vision for a welding machine for Reachbots Automation, a deep-tech startup in Singapore, in which we created a multiple of models that are retrainable which allow high precision finding of flaws in welds.  I am not going to release any data or source code, but we found an algorithm which had high accuracy, as well as laser profile detection to detect any kind of welding flaws. Here is the corporations website: Reachbots Automation  

  Omdena How we created a GIS dataset with no data Omdena Poland Chapter     Introduction   Satellite imagery has become a valuable tool for monitoring crop health and identifying crop types.  Copernicus, the European Union's Earth Observation Programme, provides free and open access to satellite data,  making it an ideal source for creating a dataset for training an AI model to provide information on crop type  and health for farmers in Poland. Creating a dataset for training an AI model involves collecting and labeling data.  In this case, we will collect satellite images of agricultural fields in Poland and label them with information on crop  type from the LUCAS Land Use and Coverage frame Survey. Contrary to most GIS datasets, which provide raster at a regional or large area level, when it comes to identifying  land usage, we understand that the details matter.  That's why we work at the field level when it comes to identifying...