# EENN Roadmap
This document outlines the development roadmap for eenn (Enlightened Equation Neural Network).
## Current Release: v0.1.0 (October 2025)
### ✅ Available Features
- **Hybrid Constraint Solving**
- Linear equation systems
- Non-linear equations (small integer domains)
- Inequality range detection
- Parentheses-aware expression parsing
- **Lightning Strike Cognitive Engine**
- Dynamic strategy selection
- Multi-backend routing (Linear, SMT, Brute-force)
- Confidence-based solution validation
- **Phase 4 Advanced Features** (NEW)
- Backend Auto-Selection with system capability detection
- Advanced Analytics with timeout rate tracking
- Z3 SMT Integration for production-grade solving
- **Optional Features**
- GPU acceleration (experimental, via wgpu 27.0)
- Zero-copy serialization (rkyv)
- Async constraint solving
### Known Limitations
- Reversed comparisons not supported (`5 < x` must be written as `x > 5`)
- Non-linear solving limited to small domains (-20 to 20 by default)
- Mixed equality/inequality constraints don't optimize ranges
---
## Next Release: v0.2.0 (Q1 2026)
### Planned Features
1. **Enhanced Parsing**
- Support reversed comparisons (`5 < x`)
- Better error messages with suggestions
- Multi-line constraint input
2. **Solver Improvements**
- Extended non-linear domain support
- Optimization for mixed constraints
- Incremental solving
3. **Neural Components**
- Expanded function registry
- More activation functions
- Basic pattern learning
4. **Developer Experience**
- Comprehensive documentation site
- More examples and tutorials
- Performance tuning guide
---
## Future Vision: v1.0.0 (2026)
### Major Goals
1. **Production-Ready Stability**
- Stable API with semantic versioning
- Comprehensive test coverage
- Battle-tested on real-world problems
2. **Advanced Neural-Symbolic Integration**
- True hybrid neural guidance
- Cross-learning between solving strategies
- Adaptive strategy improvement
3. **Extended Solver Support**
- CVC5 integration
- Additional SMT theories
- Custom theory extensions
4. **Performance**
- Optimized GPU kernels
- Parallel constraint solving
- Distributed solving support
5. **Ecosystem**
- Language bindings (Python, JavaScript)
- Integration with popular frameworks
- Cloud deployment support
---
## Long-Term Research (Post v1.0)
### Exploratory Features
These are research directions being explored:
- **Advanced Neural Functions**: Extended function registry with traditional ML operations (convolution, wavelets, transformers, etc.)
- **LLM Integration**: Using language models for constraint understanding
- **Automated Theorem Proving**: Integration with proof assistants
- **Quantum Constraint Solving**: Exploration of quantum algorithms
See [`docs/research_directions.md`](docs/research_directions.md) for detailed research plans.
---
## Contributing
We welcome contributions! Priority areas:
- **High Priority**: Bug fixes, documentation, examples
- **Medium Priority**: New solver backends, performance improvements
- **Research**: Novel neural-symbolic approaches
See [CONTRIBUTING.md](CONTRIBUTING.md) for guidelines.
---
## Feedback
Your feedback shapes this roadmap! Please:
- 🐛 Report bugs: <https://github.com/ciresnave/eenn/issues>
- 💡 Suggest features: <https://github.com/ciresnave/eenn/discussions>
- 📧 Contact: <https://github.com/ciresnave>
---
**Last Updated**: October 6, 2025