⚡ AI & Knowledge Graph Ontologies
Leveraging Large Language Models (LLMs), attention mechanisms, and Grakn/Neo4j knowledge graphs to automate static code analysis and mission after-action reports.
✨ Based in Rocket City (Huntsville, AL)
Lorin Bales
Tech Lead & Senior Data Scientist • Physics & Hardware Specialist
Pairing hardware, software, and advanced data algorithms for rapid product innovation. From AI vulnerability classification and Tactical Assault Kit (TAK) Starshield HWIL integration to embedded health sensors, autonomous drone simulation, and deep-sky astrophotography deconvolution.
🚀 Huntsville, AL
🧠
RAG Pipeline
Knowledge Graphs
❤️
Embedded Health Sensor
Over-Ear PPG & BLE
🛰️
HWIL in TAK
Starshield & TAKX Plugin
🛸
Autonomous Drone Flight
A-Star & CV LLM Reports
🌌
Deep-Sky Astrophotographer
AI Deconvolution & AI video
About Lorin Bales
Exceptional technical leadership bridging experimental physics, hardware engineering, and computational AI.
Multidisciplinary Problem Solver & Mentor
With a Bachelor of Science in Physics from the University of Alabama in Huntsville (UAH), I bring a rigorous first-principles mindset to every engineering challenge. I excel at unifying physical hardware, embedded streaming devices, and scalable cloud data pipelines with state-of-the-art AI architecture.
Whether designing 3D-printed over-ear heart rate monitors, reducing vulnerability assessments from weeks to days with C++ Abstract Syntax Tree RAG pipelines, integrating TRL 7 & 9 military hardware into the Tactical Assault Kit (TAK) over Starshield satellite C2, or capturing celestial deconvolution data from deep space, I thrive where physical reality meets computation.
Beyond technical execution, I am deeply passionate about mentoring young professionals—including leading intern engineering teams to build autonomous AI drone simulation engines from scratch—sharing project coding templates, and guiding future data scientists.
10+ Years
Defense & Commercial Systems Engineering$2.7M+
Secured Funding via Starshield & TAK C2 Demonstrations📡 Hardware-in-the-Loop & TAK C2
Integrating TRL 7 & TRL 9 hardware into TAKX/PyTAK over Starshield, 3D-printed embedded biosensors, and autonomous drone computer vision surveillance.
🔭 Deep Space Optical Deconvolution
Amateur astrophotography specializing in point spread function (PSF) correction, multi-scale AI noise reduction, and signal-to-noise optimization across deep galactic fields.
Technical Arsenal
Comprehensive proficiency spanning hardware sensors, AI systems, cloud infrastructure, and computational imaging.
🧠 AI / ML & Autonomous Systems
💻 Software & C2 Development
☁️ DB & DevOps
📡 Hardware & Embedded Systems
🌌 Deep Space Astrophotography & AI
Featured Projects
Innovative systems developed across defense, hardware-in-the-loop, autonomous flight simulation, medical sensors, and deep space imaging.
Tactical Assault Kit (TAK) & Starshield Integration
Served as Technical Lead for US Army Threat Systems Management Office (TSMO) initiatives, securing $2.7M in additional funding following successful C2 TSMO asset deployment over Starshield. Integrated TSMO hardware at TRL 7 & TRL 9 into the Tactical Assault Kit (TAK). Executed high-stakes government demonstrations of the TAKX (Java) plugin to command and control EWA and Motorola assets over Starshield via a custom PyTAK translation layer. Built an AI After Action Report engine utilizing autonomous LLM agents inside TAK Chat to synthesize mission engagements.
TAKX Java Plugin
PyTAK Translation Layer
Starshield Satellite C2
TRL 7 & TRL 9 Hardware
LLM TAK Chat Agents
EWA & Motorola Assets
Autonomous AI Drone Flight & Surveillance Engine
Served as Technical Lead for an enterprise AI Drone Simulation program guiding intern engineering teams from mission planning through autonomous flight execution. Engineered end-to-end simulation workflows where virtual drones autonomously navigate complex terrain using A* (A-Star) pathfinding, identify critical points of interest via real-time computer vision, and leverage multi-modal LLMs to maneuver, surveil, and automatically generate comprehensive post-mission reconnaissance reports.
A* (A-Star) Pathfinding
Computer Vision (CV)
Multi-Modal LLMs
Autonomous Navigation
Automated Mission Reports
Intern Engineering Lead
Vulnerability Enhancement Tool (VET)
Reduced evaluation of vulnerable software from weeks to days by pairing AI Large Language Models (LLMs) with knowledge graph ontologies. Ingests SwA static code analysis results to automatically classify vulnerabilities. Designed an LLM attention-aided RAG framework prompted directly by C++ Abstract Syntax Trees (AST) and Code Property Graphs.
LLM / RAG
Llama Index
Ollama
C++ AST
Docker Offline Container
GitLab CI/CD
Over-the-Ear PPG Biosensor System
Engineered a cost-effective, high-precision photoplethysmographic (PPG) monitor to measure cerebral blood flow dynamics. Designed a lightweight, ergonomic 3D-printed over-ear form factor enclosing continuous Bluetooth Low Energy (BLE) streaming on a custom minimal embedded Arduino device powered by lithium-ion battery. Paired with a bespoke Kotlin Android app displaying real-time PPG signals.
Arduino Hardware
BLE Streaming
Kotlin Android App
3D Printing
Biosignal Processing
'Maverick' (MAVRC) HWIL Analytics
Led development of MAVRC, a same-day dashboard for comprehensive analysis of Digital and Hardware-in-the-Loop (HWIL) evaluations across the MDS kill chain. Built robust ETL processes from AWS S3 Elasticsearch into an OOP analysis framework featuring Bokeh visualization. Researched and implemented Fréchet distance minimization and DBSCAN clustering to pair trajectories via open-source propagation.
Python / Bokeh
AWS S3 Elasticsearch
Fréchet Distance
DBSCAN Clustering
ATOMS/GFC Parsers
Object Pattern EXtractor (OPEX)
Led the OPEX team solving single-sensor 4-parameter motion solution ambiguity using Convolutional Neural Networks (CNN) to drastically reduce target uncertainty. Analyzed Measures of Performance against DIHD truth trajectories for Integrated Performance Reviews. Demonstrated Grakn AI knowledge graph integration coupled with CNNs for contextual scene-based reasoning across complex engagements.
TensorFlow CNN
Grakn AI Knowledge Graph
Trajectory Analysis
Scene-Based Reasoning
GMD Sensor Analysis & HDF Browser
Reverse engineered object classification for GMD sensors in Python, delivering the complete analysis product and briefings to government Tech Fellows. Developed a reusable, high-performance file browser interface for viewing and plotting multi-dimensional data in HDF files. Refactored legacy tools to Python, creating standardized parsers running 5x faster than previous implementations.
Python Refactoring
HDF Data Browser
Object Classification
Lockheed Martin / UEWR
Deep Space Astrophotography & AI
Capturing the cosmos: where precision optical tracking meets computational AI deconvolution.
🌌 Astrophotography Specialist
The Art of Celestial Signal Processing
With my ongoing astronomical pursuits, I honed deep-sky object acquisition techniques using motorized equatorial tracking mounts to ensure sub-arcsecond celestial targeting precision.
Because raw deep-sky photons are buried under atmospheric noise and optical distortion, my post-processing workflow treats astrophotography as a rigorous physical signal-to-noise optimization problem:
- AI Deconvolution & PSF: Computational point spread function (PSF) correction to sharpen fine stellar structures and resolve dense nebular cores.
- Multi-Scale AI Noise Reduction: Leveraging neural networks (such as NoiseXterminator) to separate faint photon signal from thermal sensor read noise without artifacting.
- Selective Stellar Signature Removal: Computational star removal algorithms to isolate and enhance delicate hydrogen-alpha and oxygen-III nebulosity in final compositions.
Professional Experience
A proven track record of technical leadership across defense, aerospace, autonomous systems, and commercial AI.
Tech Lead & Senior Data Scientist
Nov 2023 – PresentCOLSA • Huntsville, AL
Leading AI LLM prompt engineering and knowledge graph ontologies for VET (Vet Enhancement Tool). Ingesting SwA static analysis results from Cyber teams to automatically classify vulnerabilities, reducing analysis time from weeks to days. Developed C++ AST attention-guided RAG frameworks and delivered self-contained offline Docker CI/CD pipelines.
Technical Lead • US Army TSMO TAK & Starshield Integration
Jan 2024 – PresentUS Army Threat Systems Management Office (TSMO) Support
Served as Technical Lead for US Army TSMO programs, securing $2.7M in additional funding following successful C2 TSMO asset deployment over Starshield. Led integration of TSMO hardware at TRL 7 & TRL 9 directly into the Tactical Assault Kit (TAK). Executed government demonstrations of TAKX (Java) plugins controlling EWA and Motorola assets via a custom PyTAK translation layer. Developed an automated AI After Action Report engine utilizing autonomous LLM agents inside TAK Chat.
Technical Lead • Autonomous AI Drone Simulation Program
May 2024 – PresentAutonomous Systems Engineering
Served as Technical Lead for an AI Drone Simulation initiative, mentoring and guiding intern engineering teams from mission planning through autonomous flight execution. Engineered simulation pipelines where autonomous drones navigate via A* (A-Star) pathfinding, identify critical points of interest using real-time computer vision, and leverage multi-modal LLMs to maneuver, surveil, and automatically generate structured post-mission reports.
Tech Lead & Senior Software Developer
Feb 2022 – Nov 2023Torch Technologies • Huntsville, AL
Spearheaded development of 'Maverick' (MAVRC), a same-day analytics dashboard evaluating the MDS kill chain. Created automated ETL pipelines from AWS S3 Elasticsearch into object-oriented Python frameworks with Bokeh plotting. Optimized Fréchet distance metrics and DBSCAN clustering algorithms to pair and analyze complex flight trajectories.
Independent Optical & AI Researcher (Sabbatical)
Jul 2019 – Feb 2022Astrophotography & Computational Imaging
Dedicated sabbatical focused on mastering high-precision optical astrophotography, computational point spread function deconvolution, multi-scale noise reduction, and advanced stellar processing techniques.
AI Software Engineer
Apr 2018 – Dec 2019Modern Technology Solutions Inc. (MTSI) • Huntsville, AL
Led the Object Pattern EXtractor (OPEX) team resolving 4-parameter motion solution ambiguities using deep Convolutional Neural Networks. Evaluated and briefed OPEX Measures of Performance against DIHD truth trajectories for executive reviews. Integrated Grakn AI knowledge base graphs with CNNs to enable contextual scene reasoning.
Software Engineer & Honorary Junior Tech Fellow
Feb 2015 – Apr 2018deciBel Research Inc. • Huntsville, AL
Reverse engineered object classification algorithms for GMD sensors in Python, briefing senior government Tech Fellows. Supported Lockheed Martin Aegis and UEWR CD BIT contracts at COLSA Advanced Research Center. Developed custom HDF multi-dimensional data file browsers and refactored legacy parsers for a 5x processing speedup. Selected as Honorary Junior Tech Fellow (2017–2018) across cross-functional subject matter experts.
System Analyst
Jan 2014 – Feb 2015nou Systems Inc. • Huntsville, AL
Developed comprehensive analysis toolsets and plots for AN/TPY-2 (FBM) Focus Search Plan (FSP) and Threat Acquisition evaluations. Created post-mission trajectory linkages and 3D globe search plans for high-level executive briefings (GTI-04e Part 2).
System Engineer
Jan 2013 – Jan 2014Northrop Grumman • Huntsville, AL
Designed reliable data collection software extracting device logs from Motorola radios and network interfaces. Created robust MD5 checksum data management algorithms, automating a 4-hour weekly manual procedure into a 4-minute execution.
Bachelor of Science in Physics & Research Assistant
May 2011 – Spring 2013University of Alabama in Huntsville (UAH) & CSPAR
Education: B.S. in Physics (Graduated Spring 2013).
Research Assistant (CSPAR): Developed optimized searching algorithms for
current sheet structures within solar wind plasma. Analyzed ~11,400 structures across ACE,
Helios, Ulysses, and WIND spacecraft, proving them as boundaries between flux tubes and
sources of intermittent electromagnetic turbulence throughout the Heliosphere.
Certificates & Credentials
Continuous learning and cutting-edge data science mastery.
Palantir Foundry & AIP Builder Foundations
Palantir Technologies
Mastery of Palantir Foundry data engineering pipelines, Ontology architecture, and Artificial Intelligence Platform (AIP) application building for enterprise data ecosystems.
Palantir Foundry: Data Science Foundations
Palantir Technologies
Accelerated data science and machine learning workflows within Palantir systems, focusing on rapid data modeling, predictive modeling, and operational deployment.
Generative AI: Governance, Policy, and Regulation
AI Governance Institute
Comprehensive mastery of enterprise ethical AI frameworks, security risks, regulatory compliance standards, and risk mitigation policies for generative AI deployments.
Machine Learning: Regression
University of Washington / Coursera
Advanced mathematical regression techniques, bias-variance tradeoffs, feature selection algorithms, ridge/Lasso regularization, and predictive linear algorithms.
Machine Learning Foundations: Case Study Approach
University of Washington / Coursera
Hands-on implementation of machine learning models across real-world case studies including sentiment analysis, clustering, deep learning architectures, and recommender systems.
Honorary Junior Tech Fellow
deciBel Research Inc.
Selected by corporate leadership to join an elite cross-functional council of subject matter experts collaborating across defense programs to solve critical, complex technical bottlenecks.
📜 Peer-Reviewed Publication
Observation of Flux-Tube Crossings in the Solar Wind
Published in The Astrophysical Journal (2013). Systematic mathematical structure function analysis of Magnetohydrodynamics across 11,400+ current sheets from interplanetary spacecraft data (ACE, Helios, Ulysses, WIND), demonstrating electromagnetic intermittent turbulence throughout the Heliosphere.
Connect & Collaborate
Open for technical leadership consulting, AI architecture discussions, astrophotography collaborations, and mentorship.
Sub-Space Communication Relay
Whether you are looking to accelerate an AI product timeline, discuss hardware sensor challenges, or share astrophotography tips under the Huntsville dark skies, feel free to reach out across any of my active networks: