Andrew Charneski
Theoretical Systems Architect & AI Research Engineer
Westerville, OH
Executive Summary
Interdisciplinary researcher and systems architect with 20+ years experience bridging theoretical innovation and practical implementation. Creator of novel agentic frameworks, neural optimization methods, and other algorithms. Recently transitioned from enterprise engineering to focus full-time on advancing AI research and development. What began as theoretical exploration through blogging evolved into practical frameworks and platforms that push the boundaries of human-AI collaboration.
Core Competencies
- Authentic Innovation: Creating genuinely novel approaches rather than incremental improvements
- JVM ML Integration: Pioneering high-performance ML implementations in production-grade languages
- AI/ML Architecture: Novel neural network designs, optimization algorithms, and training methodologies
- Open Source Development: Multiple frameworks with comprehensive documentation and testing
- Systems Architecture: Designing scalable, innovative solutions for complex problems
Notable Projects & Contributions
Cognotik Agent Platform (2022-Present)
Cognotik & IntelliJ AI Coder: Platforms for AI-augmented development
- Designed for user sovereignty and cognitive empowerment
- Implemented novel approaches to human-AI collaboration
- Created infrastructure for “agentic authoring” processes
AI Coding Assistant JetBrains Plugin: Published IntelliJ plugin with Apache 2.0 license
- Multi-model support (OpenAI, Anthropic, Google Gemini)
MindsEye Neural Network Framework (2015-Present)
Innovation: High-performance component-based deep learning library with novel optimization algorithms
- Developed Quadratic Quasi-Newton (QQN) optimization combining SGD and quasi-Newton methods
- Created Recursive Subspace Optimization for handling gradient magnitude imbalances
- Implemented multi-GPU support with sophisticated memory management
- Pioneered test-driven development for neural networks with auto-generated documentation
- Achieved performance advantages through JVM optimization and native GPU integration
- Focused on production-ready systems with enterprise-grade reliability
Theoretical Research and Exploration
Fractal Thought Engine: Repository of theoretical papers developed through AI collaboration
- 15+ articles in the Portfolio section detailing independent research
- Agentic design patterns for AI systems
- Novel optimization algorithms (QQN, RSO) and memory management details of MindsEye
- Statistics and other ML research previously published on my blog
Professional Experience
Grubhub - Senior Software Engineer
Oct 2018 - July 2025 - Remote/Westerville, OH
- Platform Engineering: Optimized JVM-based applications, achieving significant performance improvements through profiling and tuning
- Open Source Contributions: Maintained and enhanced Apache Ranger (security) and Azkaban (workflow scheduling), contributing patches upstream
- AI/ML Innovation: Pioneered agentic AI systems for automated troubleshooting and operational intelligence
- Developer Productivity Tools: Architected full-stack application (React/TypeScript + Kotlin/Spring) leveraging LLMs to analyze Jenkins build failures, reducing mean time to resolution by enabling self-service diagnostics
- Support Optimization: Analyzed Slack support patterns using data-driven approaches, resulting in process improvements, enhanced documentation, and establishment of formal review workflows
- Knowledge Management: Developed Slack integration for automatic support request archival and searchable knowledge base creation
- DevOps Automation: Built sophisticated deployment orchestration tool supporting complex multi-stage deployments with automated rollback capabilities and audit trails
Technical Consulting
2012 - Oct 2018 - Seattle, WA
Expedia Inc: Led cloud migration of big data infrastructure
- Architected transition from on-premises Hadoop to AWS/Qubole
- Designed high-performance ad targeting system (TP95 <5ms)
- Created novel data analytics tools and workflow optimization systems
Amazon.com: Productionalized decision support systems
- Led design review and refactoring of prototype services
- Developed real-time DDoS detection and response systems
HBO Code Labs: Optimized large-scale web services
- Reduced CPU/memory load by 90% through architectural improvements
- Developed sophisticated code analysis tools
Plugged-In Technologies: Created cross-platform video conferencing app
- Developed Java/C application for real-time video streaming
- GWT and Android development for client applications
Publications & Open Source
- Cognotik: Agentic AI platform and Coding Assistant with 57k+ downloads, 64 GitHub stars
- Fractal Thought Engine: AI-augmented research site with 100+ papers
- MindsEye: Java-based deep learning framework with novel optimization algorithms and memory management
- Legacy Blog: Archive of theoretical research and open source projects
Technical Skills
Languages: Java (expert), Kotlin, Scala, C/C++, TypeScript, JavaScript, Python
ML/AI Technologies:
- Custom neural architectures, CuDNN, OpenCL
- Neural network optimization algorithms
- OpenAI API, Anthropic Claude, Google Gemini
Cloud & Infrastructure:
- AWS (EC2, S3, EMR, Lambda)
- Distributed Systems (Hadoop, Spark, Storm)
- GPU and Parallel Computing (CUDA, multi-GPU orchestration)
Development Practices:
- Test-Driven Development (TDD)
- Continuous Integration/Deployment (CI/CD)
- Agile methodologies
- Open source contribution and maintenance
- Static code analysis and transformation tools
Education
University of Illinois at Urbana-Champaign
Bachelor of Engineering in Physics, Minor in Mathematics