High-Performance Rust-Python Integrated Financial System

YOU NEED TO READ THE ATTACHED FILE!!! $10 Job I expect the contractor to pursue platform-like historical processing speed. If hardware, storage or API throughput creates a genuine lower bound, quantify it with benchmarks and recommend the exact infrastructure required. Do not simply label the existing runtime acceptable. ### Final deliverables * Optimized Rust-based processing core. * Python dashboard and research integration where appropriate. * Complete source code in a private GitHub repository. * Reproducible Windows and hosted builds. * One-click local launcher. * Hands-free hosted deployment configuration. * Persistent and compressed historical storage. * Automatic incremental OANDA collection. * Automatic discovery and validation cycles. * Permanent archive of validated `TRUE` strategies. * Fully populated Results page using the supplied real data. * Automated unit, integration, equivalence and end-to-end tests. * Performance benchmark report. * Architecture and metric-contract documentation. * Recovery and backup procedure. * No embedded credentials or secrets. * Clear evidence that all 200 engines and every required metric remain active. ### Required expertise The ideal contractor should have strong experience with: * Rust performance engineering. * Python optimization and Rust/Python integration. * Streaming and event-driven systems. * Tick-level financial data. * Line Break or comparable stateful market-structure engines. * Apache Arrow, Parquet, DuckDB or high-performance analytical storage. * Multiprocessing, concurrency and chronological state machines. * Quantitative strategy discovery and validation. * QuantConnect LEAN. * Profiling large data pipelines. * GitHub CI/CD. * Railway, containers, persistent volumes and object storage. * Windows packaging and one-click deployment. Please respond with: 1. Relevant examples of high-throughput financial or event-processing systems you personally built. 2. Your recommended target architecture. 3. How you will prove mathematical equivalence before replacing the existing implementation. 4. Your expected cold-run and warm-run performance on approximately 3.8 million source observations and 200 stateful engines. 5. Estimated peak memory and disk usage. 6. Your checkpoint and crash-recovery design. 7. How quickly you can produce the first real Results-page demonstration. I am not looking for cosmetic optimization, a prototype, or a temporary workaround. I need the complete system to reach strategy discoveries quickly, reliably and repeatedly while preserving the exact analytical behavior of the existing program.