Our Journey#
11 Years of Building Production AI Systems#
2014 — Foundation#
December 31, 2014: Positronic AI founded in St. Louis, Missouri.
2015 — Early Experiments#
WalkaboutOffice: Built a social web application with integrated videoconferencing for semi-retired executives pursuing second acts — early exploration of real-time collaboration technologies that would inform later platform work.
2016 — First Major AI Project#
PCGS Counterfeit Detection: Developed visual search and computer vision systems for the Professional Coin Grading Service to identify counterfeit coins. This project established our expertise in image analysis and pattern recognition using OpenCV and document databases.
2017-2023 — Deep Learning & Patent Development#
Stratus EEG Biomarker Discovery: Multi-year engagement developing novel neural network architectures for neurological signal processing. This work:
- Accelerated biomarker discovery from months to days
- Resulted in 5 patents for novel deep learning approaches
- Demonstrated Level 7 AI maturity—creating genuinely new model architectures
- Supported research into epilepsy, Parkinson's, and other neurological conditions
2018-2019 — Healthcare AI#
Bed Exit Prediction: Developed CNN+LSTM models analyzing infrared video to predict patient bed exits before they occur—enabling proactive intervention and reducing fall injuries in healthcare settings.
2020-2024 — Financial AI & Platform Development#
Price Movement Prediction: Built sophisticated models for securities price movement prediction, applying deep learning expertise to financial markets.
LIT Platform Development: Created an enterprise AI development platform enabling teams to build agentic applications with minimal backend code. The platform combines:
- Generative AI stack (LLM orchestration, MCP tools, workflow automation)
- Predictive AI stack (neural network design, experiment tracking, model deployment)
2024 — Investment & Pivot#
Received $1M investment to create technology offerings and patents for future licensing considerations.
November 2024: Filed patent application for LLM-driven hyperparameter tuning of machine learning training processes—a novel approach to automating one of the most expertise-intensive aspects of ML development.
2025 — Current Focus#
Fractional CTO Services: Working with select clients who are building AI-powered products, providing strategic technical leadership and hands-on development support.
Technology Licensing: Converting provisional patent to full patent application, developing licensing strategy for Spring 2025.
Methodology Development: Refining "Vibe Coding" approach—senior engineers providing intent while AI handles implementation—achieving 2-3 week sprints in single days.
The Thread Through Everything#
Looking back, a pattern emerges: every successful engagement combined deep technical expertise with genuine business impact. The value was never in the software itself—it was in knowing what to build and how to build it for each specific context.
This remains our thesis: as AI makes software creation increasingly accessible, the differentiator becomes human judgment, domain expertise, and the ability to guide organizations through fundamental change.
Ready to write the next chapter together? Contact us to discuss how we can help your organization succeed with AI.