Paul Prae

University of North Carolina at Charlotte

Data Science

January 1, 2015 – January 1, 2015

The University of Georgia

Bachelor of Science (BS) and Bachelor of Arts (BA), Computer Science and Cognitive Science

January 1, 2008 – January 1, 2012

Georgia State University Perimeter College

Associate of Science (AS), Business Administration

January 1, 2005 – January 1, 2007

Autonomize AI

Principal AI Solutions Architect

April 1, 2026 – Present

United States · Remote

Arine

Staff AI DataOps Engineer - AI Agents × Cloud Architecture × Data Platforms

September 1, 2025 – March 1, 2026

Remote

Booz Allen Hamilton

Chief AI Architect, Senior Manager

July 1, 2024 – March 1, 2025

Atlanta, Georgia, United States · Hybrid

Mento

Executive AI and Data Science Coach

September 1, 2023 – August 1, 2024

Buford, Georgia, United States · Remote

TReNDS Center

Neuroinformatics Data Platform Architect

January 1, 2022 – September 1, 2023

Atlanta, Georgia, United States · Hybrid

Hyperbloom

Chief AI Officer, Founder

June 1, 2021 – August 1, 2025

United States · Remote

Amazon Web Services

Enterprise AI and ML Solutions Architect

August 1, 2018 – May 1, 2021

Atlanta Metropolitan Area · On-site

NeuroLex Labs

Senior AI Engineer

February 1, 2018 – July 1, 2018

Atlanta Metropolitan Area · Hybrid

Decooda

Senior AI Solutions Architect

February 1, 2018 – July 1, 2018

Atlanta Metropolitan Area · On-site

Slalom Consulting

Advanced Analytics Consultant

July 1, 2015 – January 1, 2018

Atlanta Metropolitan Area · On-site

Red Ventures

Software Engineer

March 1, 2014 – July 1, 2015

Fort Mill, South Carolina · On-site

Microsoft

Senior Support Engineer

July 1, 2012 – March 1, 2014

Charlotte Metro · On-site

Paul Prae

Digital Strategy Consultant, Web Developer

January 1, 2010 – July 1, 2012

South East United States

John Keane Studios

Engineering Intern

August 1, 2009 – May 1, 2010

Athens, GA

Damon Moon and the Whispering Drifters

Production Manager for the Summer Tour

June 1, 2009 – August 1, 2009

The United States of America

Paul Prae

Music Producer, Audio Engineer

February 1, 2005 – July 1, 2012

South East United States

AI Engineering Assistant

January 1, 2025 – Present

A world-class Artificial Intelligence (AI) researcher and engineer who specializes in prompt engineering and context engineering. This assistant helps write prompts and architect agentic systems that can autonomously perform tasks, interact with tools, and pursue complex goals. These systems, capable of both planning & execution, augment human capabilities. How it works: Load prompt_engineering_assistant.system.prompt.md into Cursor or GitHub Copilot, and you get an AI expert that helps you write better prompts. Why it's useful: Instead of guessing how to structure prompts, you get step-by-step guidance, built-in validation, and copy-paste-ready results. Bonus: It can improve itself and other prompts automatically by researching the latest AI techniques.

Financial AI Assistant for Working-Class Families

June 1, 2024 – Present

This GitHub repository features an AI-driven personal financial assistant focused on budgeting, debt management, savings, and investment planning. It offers tailored financial advice and analyzes financial documents to provide actionable insights, making it ideal for users who need financial guidance or are interested in economic data analysis.

My Open Source and Local-First AI Tech Stack

January 1, 2024 – Present

I am building an AI agent/app to help people develop a career strategy and apply for jobs. I’m obsessed with open source, self-hosted, local first, multi-agent workflows. The tech stack is centered around libraries like Ollama, LangChain, n8n, and Neo4j. I use Rust and TypeScript as my programming languages. I’m focused on getting people off big tech cloud providers for benefits related to cost, privacy, security, intellectual property, control, and simplicity. The only hosting providers I plan on using are highly principled companies like Hugging Face, Mistral.ai, Together.ai, and Storj.io when local computing or storage doesn’t fit the use case. The app doesn’t need an internet connection, so I focus on local desktop apps for Mac, Windows, and Linux. I specialize in local-first app frameworks like Tauri, so I never have to deploy this app to multi-tier cloud environments, which lets me keep it hella simple and private/secure.

Career Coach and Digital Marketing AI Assistant

January 1, 2024 – Present

A career coach and digital marketing assistant designed to help job seekers create personalized go-to-market strategies and outreach content for building professional networks and securing job opportunities. The assistant crafts tailored connection requests and introductory messages by analyzing company needs.

Blockchain-Based Data Platform Architecture for Managing Life Science Data

June 1, 2022 – October 1, 2022

Applied blockchain technology to securely and privately exchange data. We designed a system to facilitate and verify a data lifecycle using smart contracts. The lifecycle identified, tracked, and secured data with event logging in place. We designed it to meet global regulatory requirements, data protection laws (CCPA, GDPR, and HIPAA), and privacy frameworks (like FAIR). We made sure it was interoperable by applying standards such as CDISC, SDTM, ADaM, and HL7.

Disaster Recovery Planning and Data Protection Solution Architecture for Clinical Trials

June 1, 2022 – December 1, 2022

Architected disaster recovery solutions and governed data for clinical trial software that helped more than 10,000 research sites in 45 countries manage their documents, data, and workflows. The platform also provides remote access so sponsors and contract research organizations (CROs) can collaborate with their sites around the world. The platform's users performed 5.5 million remote monitoring activities each month.

COINSTAC: Neuroinformatics Data Platform

January 1, 2022 – January 1, 2024

An open-source software suite that fosters collaborative research using neuroimaging and genomics data. It removes large barriers to traditional data-centric collaboration approaches. It enables groups of users to run common analyses on their own machines over their own datasets with ease. The results of these analyses are synchronized to the cloud, and undergo aggregate analyses processes using all contributor data. Decentralized pipelines allow for distributed, iterative, and feature rich analyses to be run, opening new and exciting capabilities for collaborative computation. It also offers data anonymity through differential privacy algorithms, so members do not need to fear PHI traceback.

Genomics Data Lake

June 1, 2021 – November 1, 2021

A genomics client of ours was operating entirely on-premises. Data was being stored locally on SAN disks and analytics were performed via Jupyter Hub hosted on their over-utilized HPCs. We wrote and deployed a software agent on their HPCs as a file gateway, allowing them to essentially mount a file storage system with low latency access to their newly created S3 buckets. Since the amount of data they had to move at the start was not too large, they requested for us to implement a site-to-site VPN for data transfer to save costs. We used AWS Glue to transform the data into a columnar format, Parquet, as it was optimized for the type of queries they planned to execute. To save both time and money, a transformation must occur when the technician attempts to query the data. To facilitate this, when the query is ran, a CloudWatch event is fired that causes AWS Glue to transform the VCF files into Parquet and save them into another S3 bucket in their Data Lake. The AWS Glue Data Catalog was then updated to reflect these changes. Overall, we successfully implemented a production-ready start to their data lake and automated several tasks for their data engineers.

Microservices for Web Crawling and Scraping

February 1, 2018 – May 1, 2018

Using Python and Kafka, I built a series of highly scalable microservices that could extract data from almost any popular news site. This unstructured data was then cleaned so it could be easily consumed by our text analytics platform. This system allowed our clients to follow industry trends and gain insights about both their customers and competitors. For Principal, the data was piped into an engine that predicted earnings surprise and will eventually influence investment decisions.

NeuroLex: Voice Computing for Healthcare

January 1, 2018 – May 1, 2020

We architected and developed multiple software applications, web services, and data pipelines, supporting our mission to make voice computing accessible to everyone. The automated machine learning workflows we built collected data, cleaned data, trained models, and deployed models.

Knowledge Transfer Module

January 1, 2018 – August 1, 2018

Designed, architected, and developed a system that creates high-quality training data at enterprise scale. It helps train, test, and tune machine learning algorithms for everything from search relevance and sentiment analysis to conversational agents. By annotating data with this system, business analysts were able to contribute to our knowledge base and improve the quality of predictions made by Decooda’s machine learning classifiers. The application was built with React, Express, and Node.js.

Demand Forecasting of Healthcare Services

August 1, 2017 – January 1, 2018

In order to reduce costs, optimize staffing, and support the budgeting process, we predicted demand for behavioral health crisis services in dozens of facilities across the state of Georgia. I led the data engineering and data storytelling parts of the project. I integrated data into a data warehouse from several previously siloed systems. While performing our analysis for this project, we also helped the organization improve their data governance and data science process.

Solution Assessment Comparing the Azure Bot Service to Amazon Lex for a Fortune 100 Insurance Company

July 1, 2017 – August 1, 2017

In order to properly compare the Microsoft bot ecosystem to that of Amazon’s, I designed an identical conversational interface for both. The purpose of the bots was to automate common IT support tasks. Though the flow of the conversation was the same for both bots, the back-end implementation was completely different. I led the conversational interface design for the entire project and the development of the Amazon Lex bot. Along the way, I also delivered learning sessions to their team about the bot development life cycle.

Azure as a Chatbot Service: From Purpose To Production With A Cloud Bot Architecture

January 1, 2017 – April 1, 2017

The tooling for building chatbots has exploded. Putting chatbots into production is now easier than ever. In this presentation, I focus on how you can use Azure Bot Service, Azure Search, and DocumentDB to create a scalable backend for your chatbot. By using a fully managed, serverless architecture with continuous deployment, you can get your chatbot up and running quickly. Check out this deck to learn how to combine cloud computing and artificial intelligence so you can help humans and machines achieve more together.

Neona: A Chatbot for Learning AI

October 1, 2016 – December 1, 2020

Neona is a chat bot that helps people understand artificial intelligence. By interacting with her, you can explore concepts from across the field of artificial intelligence. By studying her design and architecture, you can learn how to build conversational agents.

AI Everywhere: How Microsoft is Democratizing AI

October 1, 2016 – December 1, 2016

Microsoft has set a goal of democratizing AI, making it accessible and valuable to everyone. They're focused on building an AI stack spanning infrastructure, services, apps and agents. In this presentation, we learn how Microsoft's intelligent cloud solutions can help any organization become more proactive and differentiate themselves from intensifying competition. We will discuss: + Microsoft's AI Strategy + Examples of AI in the Enterprise + The Cortana Intelligence Suite

Predicting the Future with Azure Machine Learning

April 1, 2016 – June 1, 2016

In this talk, I focus on supervised learning, a machine learning technique for performing predictive analytics. After introducing some vocabulary, I discuss the relationship between predictive analytics and machine learning. Next, I describe how you could use a classifier, such as a decision tree, to predict which passengers survived the sinking of the Titanic. Once the machine learning process is clear, I then talk about how Azure Machine Learning is an end-to-end data science solution. Finally, I demo an experiment that predicts the outcomes of patients who went through substance abuse treatment.

An Introduction to Azure Machine Learning

March 1, 2016 – May 1, 2016

In this talk, we presented an overview of Azure Machine Learning, a fully managed cloud service that enables you to easily build, deploy, and share predictive analytics solutions. We started with the basics of machine learning and ended with a demo that used real world data.

Behavioral Healthcare Service Review Application

July 1, 2015 – January 1, 2018

Built a Universal Windows Platform app using XAML, C#, Prism, SQLite, and Azure Mobile Services and an accompanying web app using ASP.NET, Knockout.js, Bootstrap, and Power BI. I led the front-end development of both applications. The purpose of these apps is to provide nurses and behavioral staff an easy way to collect data so they can improve the quality of care patients are receiving. This data is then exposed to leadership to support strategy and operations. The project brought in over two million dollars in revenue.

IBM Watson Hackathon

May 1, 2015 – May 1, 2015

We'd like to get out a little more, intellectually speaking. To make that happen, we wanted to show Red Ventures the value of cognitive computing by performing research at IBM's Watson Hackathon (http://www.ibm.com/smarterplanet/us/en/ibmwatson/watson-hackathon.html). We aimed to augment and optimize conversations between customers and sales professionals using natural language processing and the Watson Developer Cloud. More specifically, we set out to explore chat data and dig up insights. Moving forward, we'll turn those insights into tools and recommendations for our sales teams.

HACKATHON clt: Team HTCoDelivery

March 1, 2015 – March 1, 2015

+ First place submission for the Code category for HACKATHON clt (http://www.hackathonclt.org/). + Engineered a Meteor application for a crowdsourced delivery system for Harris Teeter customers. + Provided a data-backed business launch plan by implementing a map-reduce algorithm in combination with a k-means implementation to determine neighborhoods of most active shoppers. + Processed 140+ million rows of purchase data using Apache Spark.

Predicting the Features of a Film that Maximize ROI

January 1, 2015 – May 1, 2015

For a course titled, "Big Data Analytics For Competitive Advantage", we researched trends in the film industry and applied predictive analytics to determine which types of movies would yield the highest return on investment. We performed web mining, data processing, data analysis, and data visualization to make our final prediction. We used technologies like Python, R, Excel, and SAS in our work. Overall, this course provides an introduction to the use of big data analytics as a strategic resource in creating competitive advantage for businesses. A focus is placed on integrating the knowledge of analytics tools with an understanding of how companies could leverage data analytics to gain strategic advantage. An emphasis is placed on developing the ability to think critically about complex problems/questions in real world data science and business analytics challenges.

Charlotte Startup Weekend Six: Team 'FitBloc'

January 1, 2014 – January 1, 2014

For the Sixth Charlotte Startup Weekend my team came up with the business model and wireframe behind FitBloc. The two original team members were skilled, driven, and full of energy. As the visionary, Ryan Cates pitched the idea. His passion, and the development background behind his partner, sold me immediately. The first problem we addressed: health information is increasingly difficult for the fitness community to navigate. The second problem: fitness professionals do not have a single tool-set for online marketing and branding. FitBloc is an online platform that brings the right information and resources to the right people when they need it. - https://twitter.com/fitbloc - https://www.facebook.com/fitbloc More about the event: "Startup Weekends are weekend-long, hands-on experiences where entrepreneurs and aspiring entrepreneurs can find out if startup ideas are viable. On average, half of Startup Weekend’s attendees have technical or design backgrounds, the other half have business backgrounds. Beginning with open mic pitches on Friday, attendees bring their best ideas and inspire others to join their team. Over Saturday and Sunday teams focus on customer development, validating their ideas, practicing LEAN Startup Methodologies and building a minimal viable product. On Sunday evening teams demo their prototypes and receive valuable feedback from a panel of experts." - http://startupweekend.org/

Student Studio Online

March 1, 2013 – August 1, 2013

+ Architected a JavaScript-centric web application using Node.js, MongoDB, and Angular.js. + Built several iterations of a project-based learning and content management system. Experimented with various distributions of Drupal and Moodle. + Student Studio is building a solution that is out to help urban youth develop literacy, math, and science skills through music production and business development. + We set out to integrate various web applications and native apps into a system capable of providing a social network, course management and consumption, educational gaming, virtual worlds, collaborative project management, digital media production, portfolio management, marketing, and more.‎

eChurchGive Tithing Application

May 1, 2012 – July 1, 2012

+ Led team as the project manager, designing and executing on all aspects of the application development life cycle. Collecting constant feedback from stakeholders, we developed daily iterations of an iOS application and a mobile web application. + Built the web application with jQuery, Knockout.js, and FatFractal’s NoServer. Much of the product’s back-end used a set of web and NoSQL cloud database services that were undocumented and in beta (FatFractal.com). + Contributed code to the native iPhone application. http://www.echurchgive.com/

Human Computer Interaction Course Project: Tempus Fugit

January 1, 2011 – April 1, 2011

I took this class my Senior year in College. Human Computer Interaction is one of my favorite research areas and is still central to the work I do today. Broadly stated, there are four goals for this class: 1) Understand the principles of user-centered design and how to apply them to a software-based project 2) Understand the history of human-computer interaction and how it’s changed over time 3) Explore why “good” interface design is not necessarily “common sense” 4) Utilize skills and knowledge from other disciplines in developing a software-based project During the semester we primarily focused on one project as described in the syllabus: You will undertake a group project to: + evaluate some computing-related task/problem + develop interface design alternatives for the task/problem + implement a prototype of your design + evaluate your design The theme for this semester’s projects is: Connecting to (*) You. You should think of this domain as exploring how technological innovation can help an individual become the person they want to be. Thus, the * in the domain above can be “a better”, “a wealthier”, a “healthier”, etc. By design, this is a large and rich domain to explore and we will help you investigate ideas for different problems to explore in this solution. The critical aspect of selecting a problem is that it must matter to some “real-life” people. These people can be a small group of individuals, or a large one, or any group in-between, but they will serve as your “clients” whom you must communicate with and learn from.

Logo Designs for Workfoo

July 1, 2010 – September 1, 2010

A set of logos we designed for a startup named Workfoo. Workfoo Description - The primary goal behind Workfoo is to get work done. Workfoo facilitates the creation and realization of ideas. Workfoo Slogan - Are you ready for a new economy? A freelance meritocracy is inevitable. Workfoo Mission - 100% Employment.

Widespread Panic Album Production: Dirty Side Down

January 1, 2010 – May 1, 2010

+ Shadowed Grammy-nominated producer John Keane as he engineered, mixed, mastered, and produced the album. + Prepared studio for recording sessions including microphone placement and guitar rigging. + Setup, configured, and repaired musical instruments and equipment. + Made the band comfortable.

Workfoo (backend is now FatFractal)

January 1, 2010 – March 1, 2011

My main task was to help build a web application that set out to facilitate: + Making it trivially easy to find work and get paid. + Building a support network for advancing an individual's skills and career through recommendations and experiences that are merit based. + Creating a professional freelance network that uses open source tools and its people to collaboratively get work done. + Building a perfect team that self-organizes around the technology. + Providing shrink-wrapped terms and conditions that ensure trustful and low-risk transactions. My contributions to the company depended on the current priorities. These varied among tasks such as development prioritizing, designing the user interface, testing the software, creating marketing communications, branding, building a team, designing a profit model, partnering with the college scene, and more.

AWS Certified Solutions Architect

Amazon Web Services (AWS)

June 24, 2026 – Present

AWS Certified Machine Learning - Specialty

Amazon Web Services (AWS)

June 24, 2026 – Present

AWS Certified Machine Learning Engineer

Amazon Web Services (AWS)

June 24, 2026 – Present

AWS Certified AI Practitioner

Amazon Web Services (AWS)

June 24, 2026 – Present

OTP-AWSD4: Amazon SageMaker: Simplifying Machine Learning Application Development

edX

June 24, 2026 – Present

AWS Certified Cloud Practitioner

Amazon Web Services (AWS)

June 24, 2026 – Present

edX Honor Code Certificate for Introduction to XAML and Application Development

edX

June 24, 2026 – Present

edX Verified Certificate for Data Science and Machine Learning Essentials

edX

June 24, 2026 – Present

Certificate of Completion from the Microsoft Academy for College Hires

Microsoft

June 24, 2026 – Present

Interdisciplinary Certificate in New Media

The University of Georgia

June 24, 2026 – Present

Introduction to Data Science

Coursera

June 24, 2026 – Present

Interdisciplinary Certificate in Music Business

The University of Georgia

June 24, 2026 – Present