Add ROADMAP.md: project direction and non-goals

Two clear directions:
1. "LLM의 SQLite" — embedding engine (quant.h, WASM, mobile)
2. KV compression research platform (plugin architecture, papers)

Explicit non-goals: GPU speed competition, batch serving, training.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

Author: unamedkr

ROADMAP.md

@@ -0,0 +1,80 @@
+# quant.cpp Roadmap
+
+## Vision
+
+**quant.cpp is the SQLite of LLM inference.**
+
+Not the fastest. Not the most feature-complete.
+The most embeddable, the most readable, and the only engine
+that compresses KV cache 7x without quality loss.
+
+## Positioning
+
+```
+Need speed?        → llama.cpp
+Need throughput?   → vLLM
+Need to embed LLM in your app with one file? → quant.cpp
+Need 7x longer context on the same hardware? → quant.cpp
+```
+
+## Direction 1: Embedding Engine ("LLM의 SQLite")
+
+The world's simplest way to add LLM to a C/C++ project.
+
+### Done
+- [x] quant.h single header (15K LOC, 628KB)
+- [x] 6-function API (load, new, generate, ask, free_ctx, free_model)
+- [x] WASM build (192KB binary)
+- [x] MSVC/MinGW Windows support
+- [x] Zero external dependencies
+
+### In Progress
+- [ ] API documentation (docs/api.md)
+- [ ] quant.h sync with latest source
+- [ ] Embedding examples (minimal, chat, KV compare)
+
+### Planned
+- [ ] pip install quantcpp (Python bindings)
+- [ ] iOS SDK + demo app
+- [ ] Android NDK build guide
+- [ ] Unity C# plugin
+- [ ] Unreal C++ integration
+- [ ] npm package (WASM)
+- [ ] GitHub Pages live demo with pre-loaded model
+
+## Direction 2: KV Compression Research Platform
+
+The reference implementation for KV cache quantization research.
+
+### Done
+- [x] 7 quantization types (Polar, QJL, Turbo, Uniform, TurboKV)
+- [x] Delta compression (P-frame encoding)
+- [x] QK-norm aware compression
+- [x] Plugin architecture (3 functions to add new type)
+- [x] 34 unit tests
+
+### In Progress
+- [ ] "Add Your Own Type" tutorial (docs/custom-quantization.md)
+- [ ] Arxiv tech report
+
+### Planned
+- [ ] llama.cpp KV type PR (ggml type registration)
+- [ ] vLLM KV compression plugin
+- [ ] Benchmarking suite (PPL across models × KV types)
+- [ ] Learned codebook quantization
+- [ ] Per-head adaptive bit allocation
+
+## Non-Goals
+
+- ❌ GPU speed competition with llama.cpp (requires tensor graph IR)
+- ❌ Batch serving (vLLM's domain)
+- ❌ Training support
+- ❌ 100+ model coverage
+
+## Architecture Principles
+
+1. **One file forward pass**: tq_transformer.c contains the entire inference loop
+2. **Plugin quantization**: Add types via tq_traits.c registration
+3. **Zero dependencies**: libc + pthreads only (+ Metal on macOS)
+4. **CPU-first**: NEON/AVX2 optimized, GPU as optional accelerator
+5. **Embeddable**: quant.h works anywhere a C compiler does