This issue tracks the high-level roadmap and upcoming features for Zvec. Our focus for the next major iterations is on storage scalability, mobile ecosystem maturity, and deeper integration with the AI landscape.

📱 Platform & Ecosystem Expansion

• Full Mobile Support: Complete iOS compilation support and provide idiomatic SDKs for both Android and iOS with comprehensive examples.
• Broader Language Bindings: Expand SDK capabilities to include Go, Rust, and Dart (Flutter).
• Framework Integration: Broaden support for RAG/Memory ecosystems and extend AI extensions to include more embedding models and ETL pipelines.
• Database Extensions: Explore integration with embedded databases such as SQLite and DuckDB.

💾 Storage & Resource Management

• Disk-Based Indexing: Implementation of DiskANN-inspired algorithms to support datasets exceeding RAM capacity.
• RabitQ Enhancements: Further optimization of quantization logic for better accuracy-efficiency trade-offs.
• Memory Governance: Introduce strict memory limits and constraints during the CreateIndex process to ensure system stability in resource-constrained environments.

⚙️ Feature & Performance Engineering

• Full-Text Search: Introduce keyword search capabilities alongside vector retrieval.
• Optimization Pipeline: Refine the Optimize workflow and improve write throughput for high-concurrency scenarios.
• ARM-Specific Tuning: Dedicated performance profiling and optimization for ARM-based architectures (mobile and Graviton).
• Benchmarking: Consistent participation in ANN-Benchmarks to track performance parity with industry-leading libraries.

🛡️ Stability & Quality Assurance

• Testing Coverage: Continuous expansion of API integration tests and code coverage.
• Regression Efficiency: Enhance CI/CD pipelines for faster and more reliable regression testing.

💬 We want your feedback!

This roadmap is a living document. Please head over to our Discussion Thread to suggest new features.