Eric Crook

• Hands-on builder with 24 years writing production code and 12 years leading teams.
• Designs and runs data-heavy systems — games with 2.5M DAU, analytics pipelines over 1B events/day, and LLM stacks that stay cheap.
• Back in an individual-contributor seat and happiest shipping code.

theredwizard.net/aethervanguard · Phaser, WebAudio, AWS

• Full shipped web game: hidden-reveal lane combat, roguelike narrative campaign, deckbuilder, accounts, and PvP.
• Server-authoritative backend on API Gateway + Lambda + DynamoDB; static delivery on S3/CloudFront at <$5/mo idle.
• Deploys are smoke-tested and cache-partitioned (import-graph code no-cache, heavy assets long-cache) after a real postmortem on mixed-version module graphs.
• Design system with an enforced presentation gate: no game state change ships without a visual/audio beat.

theredwizard.net/vtt · TypeScript, WebSockets, canvas

• Browser VTT built for enormous maps (up to 300×300 cells) with high visual fidelity.
• Server-authoritative fog of war: hidden geometry, tokens, and traps never reach a client that can't see them — anticheat by construction, verified in tests.
• Symmetric shadowcasting vision, dynamic torchlight, procedural dungeons, and fully synthesized audio (no sample files).

• Built an end-to-end LLM analytics stack (Python + Rust + Polars).
• Serverless architecture + Rust for dramatic cost savings.
• Added a RAG layer (LangChain + ChromaDB) for natural-language queries in Slack.
• Deployed vLLM micro-services on EKS with GPU & CPU autoscaling.
• Prompt-optimization pipeline (dSPy) + eval harness that blocks bad releases.
• Caching layer (S3 + Redis + Lambdas) to reduce LLM token cost.

• Mined Stable-Diffusion images and multi-LLM chat logs for search and health & safety policy.
• Go services for comments, search, and assorted APIs.
• Cut build-to-deploy loop time with CI/CD.
• Trimmed BigQuery dataflow, saving thousands per month.
• Trained and shipped image-categorization models on assorted GPU providers.

• Multichain Web3 solution handling NFTs across chains with OpenSea integration.
• Logic for minting, trading, destroying, and forging NFTs on-chain and in-game.

• Interactive-fiction platform letting writers turn stories into monetizing games.
• Built the analytics stack: PrestoDB, IBM Watson, QuickSight, SageMaker, Looker.
• AWS infrastructure, scaling, product updates, cost management.
• Pricing automation for better revenue.
• Handled GDPR and R&D tax credits; managed cross-zone teams.

• Full-stack work from Cordova front end to Aurora and Redshift.
• AWS architecture, scaling, and cost control.
• Random-forest models for categorization and prediction.
• Custom analytics (Redshift + Kinesis) for low-maintenance, high-throughput insights.
• GDPR compliance officer; managed engineers, analysts, and data scientists.
• Awarded 2 patents in machine learning; company publicly traded on the ASX.

• Built a big-data analytics system handling 1B+ events/day.
• Designed and led user-acquisition automation optimizing ad buys through automated experimentation.
• Managed Platform Services devops, analytics, and data-science teams across two offices for a dozen game studios.
• Built a backend for DomiNations exceeding 2M DAU.
• Recruited and led a data team serving insights to C-suite, finance, product, and QA.

• Tech lead for analytics, scalability, and cloud infrastructure.
• Built and maintained game back-ends, leaderboards, login, and security services.
• Acquired by Nexon.

• Backend work on “Millionaire City” (2.5M DAU).
• Lead backend on multiple games (300K+ DAU).
• Analytics system Tableau integrations and data engineering.

• Gloops acquisition (by Nexon).
• DomiNations mobile game — 1.5M DAU.
• Helped raise roughly $20M in startup funding.
• Machine-learning patents: 11,238,544 · 11,334,836

California Polytechnic State University, San Luis Obispo