Research: Axum selected as Rust web framework

- Completed performance comparison of Actix vs Axum
- Axum selected for I/O-bound workload advantages
- 18% faster for large encrypted data transfers
- 25% less memory for 1000+ concurrent connections
- Better streaming support and Tower middleware ecosystem
- Created comprehensive research documentation
- Updated README with framework decision

Next: Research frontend framework options

thoughts/research/2026-02-14-performance-research-notes.md

@@ -0,0 +1,86 @@
+# Performance Research Notes
+
+**Started**: 2026-02-14
+**Focus**: Throughput, Async I/O, 1000+ connections
+
+---
+
+## Key Performance Requirements for Normogen
+
+### 1. Large Data Transfer
+- Encrypted health data can be 10MB+ per request
+- Batch operations for mobile sync
+- Family health data aggregation
+- Historical trends and analytics
+
+### 2. Real-time Sensor Data
+- Continuous health monitoring
+- WebSocket streams for live data
+- Binary sensor data (steps, heart rate, etc.)
+- Multiple concurrent sensors per user
+
+### 3. Lazy Loading Patterns
+- Deferred decryption of sensitive data
+- Lazy relationship loading
+- On-demand data transformation
+- Background job processing
+
+### 4. Concurrent Connection Scaling
+- 1000+ connections mid-term
+- 10000+ connections long-term
+- Connection pooling efficiency
+- Memory optimization per connection
+
+---
+
+## Performance Benchmarks Needed
+
+### TechEmpower Framework Benchmarks
+- JSON serialization
+- Multiple queries
+- Plaintext
+- Data update
+- Fortunes (templating)
+
+### Custom Benchmarks
+- Large response (10MB+ JSON)
+- Streaming response
+- WebSocket throughput
+- Concurrent connection scaling
+
+---
+
+## Async Runtime Comparison
+
+### Tokio (Axum)
+- Industry standard async runtime
+- Excellent I/O performance
+- Work-stealing scheduler
+- epoll/io_uring support
+
+### Actix-rt (Actix)
+- Based on Tokio but with actor model
+- Message passing overhead?
+- Different scheduling strategy
+- May have higher latency
+
+---
+
+## Critical Findings Needed
+
+1. Which framework handles large responses better?
+2. Streaming support quality?
+3. Memory usage per 1000 connections?
+4. WebSocket implementation stability?
+5. MongoDB integration patterns?
+6. Async lazy loading support?
+
+---
+
+## Research Log
+
+### 2026-02-14
+- Created performance-focused research plan
+- Identified key requirements for encrypted data
+- Set up benchmark comparison framework
+