Technology
Python
Everything on this site involving Python — projects, articles, research and tools.
Does RSI Actually Work?
Testing the classic RSI(14) mean-reversion rule across 20 years and 500 instruments, with costs included and multiple-testing accounted for.
DuckDB vs pandas vs Polars on 50 Million Rows
A reproducible benchmark across six workload shapes, reporting memory as well as time — because memory is what actually stops the job.
Six RAG Chunking Strategies Compared
Fixed, recursive, semantic, structural, sentence-window and parent-document chunking measured on retrieval recall across three corpora.
AI Research & Fact-Checking System
A research agent that decomposes a claim, gathers evidence from multiple sources, cross-checks agreement between them, and refuses to answer when the evidence is thin.
RAG Evaluation Lab
Most RAG systems are tuned by vibes. This is a reproducible harness that measures retrieval and generation separately, so you can see which half is failing.
Microsoft Fabric Analytics Platform
An end-to-end Fabric implementation — lakehouse layering, incremental ingestion, dataflows, semantic model and deployment pipelines — with the decisions that are hard to reverse called out.
Laptop Data Platform
Ingestion, transformation, orchestration, testing and a BI layer — the full modern data stack, running locally on DuckDB, dbt and Dagster.
Algorithmic Trading Backtest Lab
A backtesting framework with transaction costs, slippage, walk-forward analysis and survivorship-bias handling — designed to stop a strategy looking better than it is. Educational research only.
AI Invoice Processing System
An invoice pipeline that extracts structured data, scores its own confidence per field, and routes anything uncertain to a person — because 95% accuracy on invoices is a finance problem.
Why Your RAG Evaluation Is Wrong
Most RAG evaluations measure the wrong thing, on the wrong data, with a judge that rewards the wrong behaviour. Here is what to measure instead.
Why Most Backtests Lie
Backtests fail in a consistent direction — optimistic. Here are the five structural reasons, with a detection method for each. Educational research only.
Power BI Developer to Fabric Engineer: A Roadmap
A staged plan for moving from report development into platform engineering on Microsoft Fabric — including an honest read on which existing skills carry over.
DuckDB vs pandas vs Polars: When Each One Wins
A practical decision guide based on workload shape rather than benchmark scores — because the fastest tool depends on what you are actually doing.
The Analytics Portfolio That Actually Gets Interviews
Three real projects beat thirty notebooks. What hiring managers screen for, and the structure that survives a five-minute skim.
Automation That Survives Monday Morning
Most business automations break quietly within a month. Idempotency, error handling and observability are what separate a demo from a system.
Production AI Engineering: RAG, Agents, Evals & MCP
Retrieval that holds up, agents that fail safely, evaluation that catches regressions, and MCP for tool integration.
Business Automation Engineering
Process design, idempotency, error handling and observability — the engineering discipline that separates a workflow demo from a production system.
From Power BI to Microsoft Fabric
Move from building reports to building the platform reports sit on: lakehouse architecture, medallion layering, Direct Lake and deployment pipelines.
AI-Powered Business Analyst
Requirements, analysis, documentation and stakeholder communication — with AI doing the mechanical work and you keeping the reasoning.
30-Day AI Analyst Upgrade
Thirty days, one build per week, with review. Rebuild how you work as an analyst using AI where it genuinely helps.
Algorithmic Trading Research With Python
Build a research harness that makes strategies fail honestly: point-in-time data, realistic costs, walk-forward validation. Educational only.
Modern Data Engineering on Your Laptop
DuckDB, dbt and Dagster — ingestion, transformation, testing, orchestration and lineage without a cloud bill.
Data Analyst Interview Masterclass
SQL, statistics, case studies and behavioural rounds — worked through the way an interviewer evaluates them.
AI Invoice Processing System
The complete document processing system: per-field confidence, arithmetic validation, a human review queue, and correction feedback.
RAG Evaluation System
A production-ready evaluation harness: configuration sweeps, retrieval and generation metrics measured separately, and CI regression gates.
Fabric Starter Platform
A complete Microsoft Fabric reference implementation you can adapt: notebooks, semantic model, pipelines and the architecture decision record.