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.