Quantitative research, Machine Learning systems, and data engineering, written from production experience.https://fidelius.pages.dev/https://fidelius.pages.dev/writing/industry/how-banks-work/https://fidelius.pages.dev/writing/industry/how-banks-work/A data scientist's guide to banking business models, data architecture, and regulatory constraints - built from time spent automating regulatory data pipelines across 70+ source systems at ANZ Bank.Fri, 15 Mar 2024 00:00:00 GMTbankingfinancedata-engineeringregulatoryindustryhttps://fidelius.pages.dev/writing/quantitative-research/volatility-surfaces/https://fidelius.pages.dev/writing/quantitative-research/volatility-surfaces/A practitioner's guide to implied volatility surfaces - how to construct them, what their shape encodes about market consensus, and how to extract trading edges from the information they contain.Fri, 15 Mar 2024 00:00:00 GMToptionsvolatilityderivativesmarket-microstructureivhttps://fidelius.pages.dev/writing/data-engineering/machine-learning-project-lifecycle/https://fidelius.pages.dev/writing/data-engineering/machine-learning-project-lifecycle/The real timeline of a Machine Learning project - how problem framing, data work, and deployment consistently dominate modeling time, and what this means for how to structure Machine Learning teams and projects.Fri, 01 Mar 2024 00:00:00 GMTml-engineeringproject-managementdata-sciencedeploymenthttps://fidelius.pages.dev/writing/quantitative-research/quantitative-trading-playbook/https://fidelius.pages.dev/writing/quantitative-research/quantitative-trading-playbook/A practitioner's guide to building systematic trading strategies that hold up out-of-sample - from signal research and backtesting discipline to execution modeling and live deployment.Fri, 01 Mar 2024 00:00:00 GMTtradingbacktestingsystematic-tradingrisk-managementoptionshttps://fidelius.pages.dev/writing/machine-learning/anomaly-detection/https://fidelius.pages.dev/writing/machine-learning/anomaly-detection/Statistical and Machine Learning approaches to anomaly detection - Isolation Forest, DBSCAN, autoencoders, time-series methods - and how to choose between them based on your data structure and constraints.Thu, 22 Feb 2024 00:00:00 GMTanomaly-detectionisolation-forestdbscanautoencoderstime-seriesunsupervised-learninghttps://fidelius.pages.dev/writing/industry/product-analytics-pitfalls/https://fidelius.pages.dev/writing/industry/product-analytics-pitfalls/Survivorship bias in A/B tests, Goodhart's Law in metrics, novelty effects, and the causal inference problems that make product analytics harder than it looks.Tue, 20 Feb 2024 00:00:00 GMTproduct-analyticsab-testingcausal-inferencemetricsbiasexperimentationhttps://fidelius.pages.dev/writing/data-engineering/sql-for-data-scientists/https://fidelius.pages.dev/writing/data-engineering/sql-for-data-scientists/Window functions, CTEs, time-series queries, and optimization techniques - SQL patterns that data scientists use daily but often learn inefficiently from tutorial sites that stop at basic SELECT.Sun, 18 Feb 2024 00:00:00 GMTsqlbigquerywindow-functionsctedata-scienceanalyticshttps://fidelius.pages.dev/writing/data-engineering/bigquery-vs-tensorflow-transform/https://fidelius.pages.dev/writing/data-engineering/bigquery-vs-tensorflow-transform/When to compute features in BigQuery versus TFX - the tradeoffs between SQL-native simplicity and training-serving skew prevention, based on real experience at Blue Yonder.Thu, 15 Feb 2024 00:00:00 GMTbigquerytensorflow-transformfeature-engineeringmlopspipelinestraining-serving-skewhttps://fidelius.pages.dev/writing/machine-learning/explainable-ai-in-practice/https://fidelius.pages.dev/writing/machine-learning/explainable-ai-in-practice/When model explanations actually matter and when they don't - a practitioner's guide to SHAP, LIME, attention visualization, and the hard questions about trust and accountability in machine learning systems.Thu, 15 Feb 2024 00:00:00 GMTexplainabilityshapinterpretabilityml-engineeringfairnesshttps://fidelius.pages.dev/writing/quantitative-research/options-greeks-risk-dials/https://fidelius.pages.dev/writing/quantitative-research/options-greeks-risk-dials/A working practitioner's guide to delta, gamma, theta, vega, and rho - not as formulas to memorize but as risk instruments to manage in a live options book.Thu, 15 Feb 2024 00:00:00 GMToptionsgreeksrisk-managementderivativesquanthttps://fidelius.pages.dev/writing/data-engineering/python-performance/https://fidelius.pages.dev/writing/data-engineering/python-performance/Practical techniques for making Python code fast - NumPy vectorization, Numba JIT compilation, multiprocessing, profiling tools, and common patterns that silently kill performance.Mon, 12 Feb 2024 00:00:00 GMTpythonperformancenumpynumbaprofilingoptimizationvectorizationhttps://fidelius.pages.dev/writing/quantitative-research/the-trader-mindset/https://fidelius.pages.dev/writing/quantitative-research/the-trader-mindset/What separates systematic traders from discretionary ones - market regime classification, position sizing, psychological discipline, and why most people fail at trading even when they're smart.Thu, 08 Feb 2024 00:00:00 GMTtradingsystematic-tradingposition-sizingrisk-managementpsychologymarket-regimeshttps://fidelius.pages.dev/writing/machine-learning/losses-and-metrics/https://fidelius.pages.dev/writing/machine-learning/losses-and-metrics/A practical reference covering every major loss function and evaluation metric - what each one measures, when to use it, and what it gets wrong.Mon, 05 Feb 2024 00:00:00 GMTloss-functionsmetricsevaluationmachine-learningclassificationregressionhttps://fidelius.pages.dev/writing/machine-learning/neural-network-training-playbook/https://fidelius.pages.dev/writing/machine-learning/neural-network-training-playbook/A practitioner's guide to training neural networks - from initialization and optimization to regularization, debugging, and the decisions that actually determine whether your model converges.Thu, 01 Feb 2024 00:00:00 GMTdeep-learningneural-networkstrainingoptimizationpytorchhttps://fidelius.pages.dev/writing/machine-learning/overfitting-is-a-thinking-problem/https://fidelius.pages.dev/writing/machine-learning/overfitting-is-a-thinking-problem/The bias-variance tradeoff reframed as a failure of reasoning, not tuning. Why overfitting in quantitative finance is uniquely dangerous, and how to detect and prevent it systematically.Thu, 01 Feb 2024 00:00:00 GMToverfittingbias-varianceregularizationcross-validationquantitative-financegeneralizationhttps://fidelius.pages.dev/writing/machine-learning/count-data-models/https://fidelius.pages.dev/writing/machine-learning/count-data-models/When your target variable is a non-negative integer, standard regression breaks down. A practical guide to Poisson, negative binomial, and zero-inflated models - and when each one applies.Sun, 28 Jan 2024 00:00:00 GMTcount-datapoissonnegative-binomialprobabilistic-forecastingstatisticsregressionhttps://fidelius.pages.dev/writing/machine-learning/probability-as-operating-system/https://fidelius.pages.dev/writing/machine-learning/probability-as-operating-system/Bayesian reasoning, belief updating, and calibrated uncertainty - how probabilistic thinking changes the way you interpret evidence and make decisions under uncertainty.Thu, 25 Jan 2024 00:00:00 GMTprobabilitybayesian-reasoninguncertaintydecision-makingstatisticshttps://fidelius.pages.dev/writing/machine-learning/decision-trees-and-ensembles/https://fidelius.pages.dev/writing/machine-learning/decision-trees-and-ensembles/How decision trees work, why they overfit, and how ensemble methods - bagging, boosting, and stacking - transform weak learners into the models that dominate tabular Machine Learning competitions.Mon, 22 Jan 2024 00:00:00 GMTdecision-treesrandom-forestxgboostgradient-boostingensemblesmachine-learninghttps://fidelius.pages.dev/writing/industry/data-science-in-supply-chain/https://fidelius.pages.dev/writing/industry/data-science-in-supply-chain/A practitioner's overview of how Machine Learning is applied in supply chain - from demand forecasting and inventory optimization to markdown pricing and fulfillment capacity, with what the models can and can't solve.Sat, 20 Jan 2024 00:00:00 GMTsupply-chaindemand-forecastinginventoryindustryhttps://fidelius.pages.dev/writing/machine-learning/machine-learning-taxonomy-building-blocks/https://fidelius.pages.dev/writing/machine-learning/machine-learning-taxonomy-building-blocks/A reference-first guide to the full landscape of machine learning - problem types, algorithm families, and the four universal components that every Machine Learning system shares.Sat, 20 Jan 2024 00:00:00 GMTmachine-learningsupervised-learningunsupervised-learningalgorithmsfundamentalshttps://fidelius.pages.dev/writing/quantitative-research/backtesting-framework/https://fidelius.pages.dev/writing/quantitative-research/backtesting-framework/The common mistakes that make backtests look better than reality - and the engineering disciplines that close the gap between simulated and live performance.Sat, 20 Jan 2024 00:00:00 GMTbacktestingquantsystematic-tradingrisk-managementwalk-forwardhttps://fidelius.pages.dev/writing/machine-learning/feature-engineering-deep-dive/https://fidelius.pages.dev/writing/machine-learning/feature-engineering-deep-dive/A practitioner's guide to feature engineering - the craft of transforming raw data into model-ready representations that capture what actually matters for the prediction task.Mon, 15 Jan 2024 00:00:00 GMTfeature-engineeringmachine-learningdata-sciencetime-serieshttps://fidelius.pages.dev/writing/data-engineering/data-engineering-roadmap/https://fidelius.pages.dev/writing/data-engineering/data-engineering-roadmap/A structured map of the data engineering landscape - from OS fundamentals and SQL through distributed compute, streaming, cloud services, and orchestration. Built from real project experience across Blue Yonder, Mastertrust, and independent data systems work.Wed, 10 Jan 2024 00:00:00 GMTdata-engineeringbig-datasparkkafkaairflowsqlhadoopawsarchitecturehttps://fidelius.pages.dev/writing/machine-learning/data-science-pipeline/https://fidelius.pages.dev/writing/machine-learning/data-science-pipeline/A practitioner's map of the complete data science workflow - from problem framing and data collection to deployment and monitoring - with what actually goes wrong at each stage.Mon, 01 Jan 2024 00:00:00 GMTdata-sciencepipelineml-engineeringbest-practiceshttps://fidelius.pages.dev/writing/data-engineering/tensorflow-distributed-training/https://fidelius.pages.dev/writing/data-engineering/tensorflow-distributed-training/A practical guide to TensorFlow's distribution strategies - how each works, when to use MirroredStrategy vs. ParameterServerStrategy, and the tradeoffs that determine which is faster.Fri, 10 Feb 2023 00:00:00 GMTtensorflowdistributed-trainingmlopsgpuinfrastructurehttps://fidelius.pages.dev/writing/machine-learning/image-data-deep-learning/https://fidelius.pages.dev/writing/machine-learning/image-data-deep-learning/A practical map of the deep learning landscape for image understanding - object detection vs. semantic segmentation, the key architectures for each, and which metrics to use.Wed, 01 Feb 2023 00:00:00 GMTcomputer-visiondeep-learningobject-detectionsegmentationmlhttps://fidelius.pages.dev/writing/machine-learning/differentiation-in-tensorflow/https://fidelius.pages.dev/writing/machine-learning/differentiation-in-tensorflow/How TensorFlow's automatic differentiation works under the hood, when to use GradientTape over Keras fit(), and how to build custom training loops for research and production models.Fri, 20 Jan 2023 00:00:00 GMTtensorflowautomatic-differentiationdeep-learningcustom-trainingmlhttps://fidelius.pages.dev/writing/industry/building-information-modeling/https://fidelius.pages.dev/writing/industry/building-information-modeling/How BIM creates a structured digital twin across a building's full lifecycle - and where Machine Learning applications in predictive maintenance, energy optimization, and construction quality control are emerging.Tue, 17 Jan 2023 00:00:00 GMTbimconstructiondigital-twiniotindustryhttps://fidelius.pages.dev/writing/machine-learning/scaling-in-machine-learning/https://fidelius.pages.dev/writing/machine-learning/scaling-in-machine-learning/How machine learning systems scale across three dimensions - data volume, model size, and inference throughput - and the engineering tradeoffs at each level.Sun, 15 Jan 2023 00:00:00 GMTscalingdistributed-trainingml-systemsproductioninfrastructurehttps://fidelius.pages.dev/writing/machine-learning/useful-ml-concepts/https://fidelius.pages.dev/writing/machine-learning/useful-ml-concepts/Three underrated concepts that separate production-ready Machine Learning from research prototypes - probability calibration, robust model fitting with RANSAC, and understanding all Machine Learning algorithms as variations on a single loss minimization framework.Wed, 11 Jan 2023 00:00:00 GMTcalibrationransacloss-functionsmlproductionhttps://fidelius.pages.dev/writing/productivity/learn-one-thing-at-a-time/https://fidelius.pages.dev/writing/productivity/learn-one-thing-at-a-time/A focused collection of learning resources for applied Machine Learning practitioners - from opportunity sizing frameworks to Machine Learning process courses and the embedding vs. dense layer distinction.Wed, 11 Jan 2023 00:00:00 GMTlearningresourcescareermachine-learningproduct-analyticshttps://fidelius.pages.dev/writing/machine-learning/dimensionality-reduction/https://fidelius.pages.dev/writing/machine-learning/dimensionality-reduction/A practical guide to the three main dimensionality reduction techniques - when to use each, what they preserve, and how to avoid the common mistake of using t-SNE embeddings as features.Sun, 08 Jan 2023 00:00:00 GMTdimensionality-reductionpcat-sneumapunsupervised-learninghttps://fidelius.pages.dev/writing/machine-learning/clustering/https://fidelius.pages.dev/writing/machine-learning/clustering/A technical reference for the three main clustering families - density-based (DBSCAN), centroid-based (K-Means), and hierarchical - covering their mathematical foundations, hyperparameter selection, and failure modes.Tue, 03 Jan 2023 00:00:00 GMTclusteringunsupervised-learningdbscank-meansmlhttps://fidelius.pages.dev/writing/machine-learning/support-vector-machines/https://fidelius.pages.dev/writing/machine-learning/support-vector-machines/SVMs from first principles - the margin maximization objective, soft margins, the kernel trick, and the practical cases where SVMs outperform and where they don't.Sat, 31 Dec 2022 00:00:00 GMTsvmclassificationkernel-methodsmlsupervised-learninghttps://fidelius.pages.dev/writing/industry/be-up-to-date-data-industry/https://fidelius.pages.dev/writing/industry/be-up-to-date-data-industry/A curated list of engineering blogs from top technology companies - where practitioners publish real Machine Learning system designs, infrastructure decisions, and data science case studies.Thu, 29 Dec 2022 00:00:00 GMTresourceslearningdata-scienceengineering-blogsindustryhttps://fidelius.pages.dev/writing/productivity/data-structures-algorithms/https://fidelius.pages.dev/writing/productivity/data-structures-algorithms/The core intuition behind space-time complexity analysis, with a guide to the best resources for building DSA fundamentals as a data scientist.Sat, 17 Dec 2022 00:00:00 GMTdsaalgorithmspythoncomplexityinterview-prephttps://fidelius.pages.dev/writing/productivity/research-papers-resources/https://fidelius.pages.dev/writing/productivity/research-papers-resources/A curated reference of Machine Learning research papers organized by topic, and Python libraries for graph analysis, explainable AI, trading, and MLOps - actively maintained as a working reading list.Sat, 17 Dec 2022 00:00:00 GMTresearchpaperspythonlibrariesresourcesmachine-learninghttps://fidelius.pages.dev/writing/quantitative-research/trading-system-notes/https://fidelius.pages.dev/writing/quantitative-research/trading-system-notes/A structured survey of how major asset classes behave and what drives their systematic trading opportunities - from cash equities and fixed income to derivatives, commodities, and forex.Wed, 01 Jun 2022 00:00:00 GMTtradingasset-classesalgorithmic-tradinghftmarket-makinghttps://fidelius.pages.dev/writing/data-engineering/tfx-ml-pipelines/https://fidelius.pages.dev/writing/data-engineering/tfx-ml-pipelines/A ground-up walkthrough of TensorFlow Extended - orchestrators, metadata, standard components (ExampleGen through Pusher), and building custom components. Written from hands-on work building Machine Learning pipelines in 2022.Fri, 22 Apr 2022 00:00:00 GMTtfxmlopstensorflowpipelinesml-engineeringairflowkubeflowfeature-engineeringhttps://fidelius.pages.dev/writing/productivity/markdown-tricks/https://fidelius.pages.dev/writing/productivity/markdown-tricks/Building documentation sites with MkDocs and Jupyter Book - setup commands, the plugins worth using, admonition syntax, and the badge and icon formatting that makes docs look professional.Wed, 05 Jan 2022 00:00:00 GMTmarkdownmkdocsjupyter-bookdocumentationtoolshttps://fidelius.pages.dev/writing/industry/export-import-guide/https://fidelius.pages.dev/writing/industry/export-import-guide/A practical overview of India's export-import regulatory framework - IEC, FEMA, Incoterms 2020, trade documentation, government incentive schemes, and payment instruments.Tue, 04 Jan 2022 00:00:00 GMTtradeexport-importindiasupply-chainlogisticshttps://fidelius.pages.dev/writing/productivity/pop-os-setup/https://fidelius.pages.dev/writing/productivity/pop-os-setup/A repeatable bash setup script for a Pop!_OS data science workstation - system tools, Chrome, Docker, SQLite, TA-Lib for quantitative finance, Python environment, and SSH for GitHub.Tue, 04 Jan 2022 00:00:00 GMTlinuxpop-ossetuptoolspythondockerhttps://fidelius.pages.dev/writing/productivity/obsidian-tricks/https://fidelius.pages.dev/writing/productivity/obsidian-tricks/Practical Obsidian setup for technical note-taking - the essential plugins, Dataview queries for dynamic indexes, and workflows for keeping a knowledge vault organized.Mon, 03 Jan 2022 00:00:00 GMTobsidiannote-takingtoolsproductivityknowledge-managementhttps://fidelius.pages.dev/writing/productivity/python-resources/https://fidelius.pages.dev/writing/productivity/python-resources/Curated Python libraries organized by use case - graph analysis, explainable AI, quantitative finance, web frameworks, and MLOps tooling.Mon, 03 Jan 2022 00:00:00 GMTpythonlibrariestoolsdata-sciencemlopshttps://fidelius.pages.dev/writing/machine-learning/linear-neural-networks/https://fidelius.pages.dev/writing/machine-learning/linear-neural-networks/The mathematical foundation of linear regression and logistic regression - what they optimize, what assumptions they make, and why understanding these fundamentals matters for every model built on top of them.Sun, 02 Jan 2022 00:00:00 GMTlinear-regressionlogistic-regressionloss-functionsmlfoundationshttps://fidelius.pages.dev/writing/productivity/jupyter-lab-tips/https://fidelius.pages.dev/writing/productivity/jupyter-lab-tips/Practical JupyterLab configuration for data science work - remote access over the network, useful extensions for visualization and productivity, and embedding media in notebooks.Sun, 02 Jan 2022 00:00:00 GMTjupytertoolspythondata-sciencesetup