2 min read
Twelve Iterations on a Contact Map and a Snappier Shared Inbox
A day of visual iteration on the contact map, plus performance and correctness work in the shared inbox.
FrontendDesignPerformanceEmailSEA
Read more
2 min read
Pre-Integration Guards, Faster Mail, and the Sales Shared Inbox
A 46-commit day making the integration safe to deploy, speeding up Mail, and building the sales shared-inbox surface.
LaravelPerformanceEmailIntegrationDeployment
Read more
2 min read
Candle Continuity, Local-First Market Data, and Deploy Health
A 32-commit day making candle-service fast and continuous while moving dex-trader to local-first data.
FastAPISQLitePerformanceTradingData Pipelines
Read more
2 min read
Regime Feeds, Volume Workers, and Curation in Candle Service
Adding a Binance regime feed, backfilling volume, and dropping a concept that didn't earn its complexity.
PythonMarket DataSQLitePerformanceRefactor
Read more
2 min read
Building Edge Lab and Porting It to a Triple-Arc XPU Rig
Phase 1 through 9 of a trading research platform, then a hard pivot from numba-dpex to torch.xpu, Triton, and IPEX.
PythonTritonIntel ArcPerformanceBacktesting
Read more
2 min read
OpCache, JIT, and a Slimmer Docker Build
Trimming image build time while turning on PHP performance features and extracting a reusable migration kit.
DockerPHPPerformanceOpCacheSQLite
Read more
2 min read
26 Iterations: Porting Cleanroom Kernels to Gemma 4
A single day of rapid iteration took the Gemma 4 port from a baseline to validated QKV fusion, invalidating several assumptions along the way.
SYCLKernelsLocal LLMPerformanceIteration
Read more
2 min read
All-Local 29 Tokens Per Second and Fused Norm Kernels
Fusing QKV, inlining norm-weight application, and pushing block loads through the Q4_K path.
SYCLKernelsPerformanceLocal LLMOptimization
Read more
2 min read
Standalone Inference Working: Shared Memory and a CPU Fallback
The standalone backend finally produced inference after fixing ABI layout, shared-memory buffers, and a simple CPU fallback.
SYCLKernelsABILocal LLMPerformance
Read more
2 min read
Enabling Local Dispatch Op by Op Until the Graph Ran Entirely on My Kernel
47 commits and a progressive enablement series that ended with full local decode dispatch at 7.0 t/s and 100% correctness.
SYCLKernelsPerformanceDebuggingLocal LLM
Read more
3 min read
Fixing 20-60 Second Page Loads Across the AustinsElite Codebase
A performance day: SQLite query fixes, frontend bottlenecks, and performance smoke tests to keep it from regressing.
SQLitePerformanceLaravelPlaywrightFrontend
Read more
2 min read
Writing an ESIMD Fused Kernel From Reverse-Engineered Assembly
SLM-tiled GEMM, SPIR-V decompilation, and a corrected Q4_K dequant that finally produced coherent output.
SYCLESIMDKernelsReverse EngineeringPerformance
Read more