Jul 2026First paper accepted, and RegimeFactorZoo reaches v1.0
- "Training Data Set Effects in LLM-Assisted Quantum Portfolio Optimization" accepted to QNLP-AI 2026 and selected for an oral presentation — a quantified look-ahead-bias caution for evaluating LLMs in finance.
- "Empirical L1 Cache Characterization of Quantitative Finance Kernels on AMD EPYC Zen 2" (the qhpc_cache study) submitted to ACM AIMLSystems 2026 — three cache-behavior regimes, with a 1,657× L1-miss jump at the 32 KB boundary.
- Completed RegimeFactorZoo v1.0: reconstructed the Fama-French SMB and HML factors from raw CRSP/Compustat data and matched Kenneth French's published series at ρ = 0.976 (SMB) and 0.965 (HML) across 726 months.
- Earned the IBM Generative AI Engineering Professional Certificate (16 courses), and finished MITx 15.415.1x — Foundations of Modern Finance I — with honors.
- Attended the ARPM Bootcamp onsite: buy-side estimation, projection, pricing, and portfolio/risk construction end to end.
Jun 2026Rebuilt this site and pushed RegimeFactorZoo's data layer
- Stood up a reproducible factor-model pipeline on fully public data (Kenneth French library + Open Source Asset Pricing) after losing summer WRDS access — a strictly more reproducible design.
- Validated Fama-French three-factor returns against the published series before trusting anything downstream.
May 2026Quantum foundations, the slow and careful way
- Working through Sutor and Wong toward Nielsen-Chuang: qubits as unit vectors, gates as unitaries, measurement as the one non-linear step.
- Mapped where quantum amplitude estimation offers a quadratic speedup over the classical Monte Carlo I had just been profiling.
Apr 2026Joined Prof. Yuxi Hong's lab
- Undergraduate research assistant on the linear-algebra and systems side of quantum computing.
- Weekly meetings and written progress; focus on matrix algebra and large-scale sparse computation.
Early 2026Captain Whiskers — a quantum-optimized trading agent
- Built an autonomous trading agent fusing variational-quantum portfolio optimization, Byzantine-fault-tolerant verification, and post-quantum signatures for the LabLab AI Trading Agents hackathon.
- The first build where the HPC, finance, and quantum threads met in one place.
2025 → 2026qhpc_cache — measuring what the cache actually does
- Empirical L1-cache characterization of Cholesky, Monte Carlo, GARCH, and GEMM on AMD EPYC, instrumented with PAPI on BigRed200.
- Found that three of four finance kernels are decided by memory behavior, not flop count — and tied every claim to a hardware counter.
- Written up as a paper, now under review at ACM AIMLSystems 2026.
2025Tooling and applied AI
- Shipped a bounded-memory filesystem-audit toolkit and a Chrome MV3 ethics-review extension.
- Completed the IBM Generative AI Engineering certificate and applied RAG techniques in a document-diligence pipeline.