Jan 7, 2026 · 8 min read
ML in Finance: A Decision System, Not a Forecasting Contest
Why financial ML must be designed around decisions, uncertainty, and controls—not a single accuracy score.
MLFinanceRisk
Read more →Writing
Research notes on AI systems that matter.
Twelve research-backed essays on financial ML, agentic systems, Apex Arena, and model adaptation.
Jan 28, 2026 · 7 min read
Credit Models Need Reasons, Not Just Rankings
A practical framework for pairing predictive credit models with explanation, fairness review, and human judgment.
Credit RiskExplainabilityML
Read more →Feb 18, 2026 · 8 min read
Backtesting Is an Argument About the Past
How leakage, costs, and regime change turn a promising financial backtest into a fragile production strategy.
QuantEvaluationML
Read more →Mar 4, 2026 · 9 min read
Agentic AI in Medicine: Autonomy Must Stop at the Right Boundary
A safety-first view of agentic workflows for clinical support, where tools, approvals, and audit trails define usefulness.
Agentic AIHealthcareSafety
Read more →Mar 25, 2026 · 8 min read
A Safer Pattern for Financial Agents: Propose, Verify, Approve, Execute
Why financial agents should be built as controlled workflows with segregated duties and explicit approval gates.
Agentic AIFinanceGovernance
Read more →Apr 15, 2026 · 8 min read
Evaluating Agents Means Testing the Whole Loop
A test strategy for agentic systems that measures tool use, abstention, recovery, and reviewer experience—not only final answers.
Agentic AIEvaluationReliability
Read more →May 6, 2026 · 11 min read
Apex Arena: Building Evidence-Linked Race Rooms
How Apex Arena turns live Formula racing data into a bounded, evidence-linked conversation between five specialist agents.
Apex ArenaFormula RacingAgentic AI
Read more →Jun 3, 2026 · 12 min read
Hybrid Token-Efficient Routing: Spend Tokens Only When a Proof Fails
A separate project: a Python routing agent that uses deterministic solvers first, then gated local inference, then a cloud fallback.
Hybrid RoutingToken EfficiencyPython
Read more →Jul 22, 2026 · 12 min read
Completing Apex Arena: Shipping the Live Data and Replay Stack
Apex Arena is complete: a production Formula racing experience with resilient data handling, replayable race rooms, and a verified delivery path.
Apex ArenaLive DataFull Stack
Read more →May 20, 2026 · 8 min read
Fine-Tuning Starts With a Data Contract
Before choosing LoRA or full tuning, define the task, data lineage, quality thresholds, and non-goals.
Fine-TuningDataLLMs
Read more →Jun 17, 2026 · 8 min read
Choosing Full Fine-Tuning, Adapters, or LoRA
A decision guide for selecting an adaptation strategy based on task shift, cost, inference constraints, and operational needs.
Fine-TuningLoRAPEFT
Read more →Jul 8, 2026 · 8 min read
Fine-Tuning Is a Lifecycle, Not a Training Run
How to operate an adapted model with evaluation gates, versioning, monitoring, and a disciplined rollback path.
Fine-TuningMLOpsGovernance
Read more →