Strategic consulting and enterprise solutions architecture
Andrew Charneski
Software Consultant & Systems Architect
Executive Summary
Strategic technology consultant and systems architect with 20+ years of experience advising Fortune 500 clients (Expedia, Amazon, HBO) on complex technical challenges. Specializes in AI implementation strategy, enterprise architecture, and performance optimization, delivering transformative results including 90% efficiency improvements and successful cloud migrations. Bridges the gap between cutting-edge research and practical business solutions, helping organizations leverage technology for a competitive advantage.
Core Competencies
Strategic Technology Advisory: Guiding stakeholders and technical teams on strategy, digital transformation, and large-scale system architecture.
Enterprise Solution Architecture: Designing and delivering innovative, scalable solutions for big data, distributed systems, and AI, with a focus on business value and ROI.
Performance & Cost Optimization: Proven track record of identifying and resolving critical performance bottlenecks, achieving up to 90% load reduction and significant cost savings for clients.
Innovation to Production: Translating novel concepts from AI research and optimization theory into practical, production-ready enterprise systems.
JVM & Cloud Expertise: Deep, hands-on expertise in JVM optimization, cloud migration (AWS), and building high-performance, resilient infrastructure.
Professional Experience
AI Strategy & Implementation: Prototyped various agentic AI systems to provide an operational intelligence platform, reducing manual intervention and support costs.
Solution Architecture: Designed a full-stack solution using LLMs to diagnose complex build failures, establishing a new capability for developer self-service and reducing mean time to resolution.
Performance Optimization: Conducted deep performance analysis of mission-critical JVM applications, delivering significant improvements and providing a framework for continuous optimization.
Open Source Strategy: Guided the adoption and enhancement of key open-source technologies (Apache Ranger, Azkaban), ensuring alignment with enterprise security and workflow requirements.
DevOps Transformation: Architected a modern deployment orchestration platform to de-risk complex releases, enabling safer and faster delivery cycles for multiple engineering teams.
Expedia Inc (Oct 2014 - Oct 2016; April 2017 - Oct 2018): Led cloud migration of big data infrastructure to AWS/Qubole and architected a high-performance real-time ad targeting system (TP95 <5ms, ~10k tps).
Amazon.com (Nov 2016 - Feb 2017): Led the productionalization of a prototype Java web service for decision support and automation.
HBO Code Labs (Dec 2013 - Sep 2014): Reduced CPU/memory load by 90% on large-scale web services and developed sophisticated static code analysis tools.
Various (April 2011 - Nov 2013): Created a cross-platform video conferencing app (Plugged-In Technologies) and an Android video game streaming client (Big Fish Games).
Distributed Energy Management (2010-2011): As Team Lead and Architect, led a team of six, designed a high-performance data service and analytics platform for time series data.
Amazon.com (2007-2009): As SDE II, created a distributed real-time data reporting service and developed Apache httpd C modules for routing and security.
Aristocrat Technologies, Inc (2005-2007): As a Software Engineer, developed C# .NET commercial business applications for the gaming industry.
Early Consulting/Internships (1998-2004): Roles at Texas Instruments, Lucent Technologies, and others, working with C++, Java, and web technologies.
Publications & Open Source
Quadratic Quasi-Newton and Recursive Subspace Optimization
Algorithm ResearchBreakthrough optimization algorithms bridging first/second-order methods through quadratic interpolation. RSO treats per-layer learning rates as optimization problem with emergent regularization
MindsEye: Deterministic GPU Memory Management for Deep Learning
Open Source FrameworkComprehensive neural network library with reference counting for GPU resources, object pooling, and memory pressure response. Supports CuDNN, OpenCL, Spark clusters. Anticipated memory challenges still plaguing modern frameworks
Geometric Symmetry in Deep Texture Generation
Research ImplementationBreakthrough in neural art achieving perfect mathematical symmetry through kaleidoscopic preprocessing as part of the DeepArtist platform. Works across Euclidean, spherical, and hyperbolic geometries.
Cognotik: Cognitive Pluralism in AI Systems
SaaS SolutionAI platform implementing four distinct cognitive modes (TaskChat, PlanAhead, AutoPlan, GoalOriented) embodying different philosophical frameworks. Demonstrates measurable behavioral differences based on philosophical foundations
reSTM: Distributed Software Transactional Memory
Distributed Systems ResearchREST-based transactional memory providing ACID guarantees across distributed clusters. Implements MVCC with fine-grained locking, actor-based architecture, and formal correctness proofs. Near-linear scaling properties
Technical Skills
Consulting
Technical
Languages
Practices
Education
University of Illinois at Urbana-Champaign
Bachelor of Engineering in Physics
Minor in Mathematics