Work

Evidence library

Demos plus repo-backed implementations—browse featured work first, then the rest of the catalogue.

Featured systems

Short list of the strongest references—prep tooling, KPI-style apps, scoring, HTTP services. Click a card or open GitHub/live links.

Featured · Interactive demo

Data cleaning toolkit

Streamlit
pandas
pytest
Parquet / JSON

Upstream prerequisite, not a side utility: downstream models and dashboards ingest the same reviewed tables—multi-format inputs become auditable CSV/Parquet/JSON plus HTML step log and before/after views, capped near 100K rows; rules cover bad formats, duplicates, skewed categories, optional outliers, plus bundled samples for dry runs.

GitHubLive demo

Featured · Interactive demo

KPI dashboard implementation example

Streamlit
Sales & marketing CSV
Mapping presets
Optional OpenAI

An interactive implementation reference—not a SaaS subscription: preset mappings and hard row gates produce KPI cards, trends, breakdowns, optional deltas, and a rule-first “what changed?” so stakeholders see how a scoped Sheets or BI dashboard engagement would behave.

GitHubLive demo

Featured · ML systems reference

Batch scoring pipeline

Python
Batch
pandas
pytest

Scores many records in one run on a schedule, with the same rules each time instead of one-off checks.

GitHubLive demo

Featured · ML systems reference

Customer churn prediction system

FastAPI
scikit-learn
Classification
Joblib

Shows a production-style pattern: one pinned model artefact, fixed probability and tier outputs, and an HTTP surface your stack can call—without silent notebook drift on the live path.

GitHubLive demo

More interactive demos

Remaining Streamlit apps—validation, exports, and safe operator limits (featured demos are above).

Interactive demo

EDA report generator

Streamlit
pandas
pytest
Jinja2

Read-only dossier you can archive or forward—no cell edits: capped sampling with sheet picker, full column intelligence, correlations, histograms, and warnings, rendered to HTML from memory with optional PDF when WeasyPrint is available.

GitHubLive demo

Interactive demo

AI-assisted data analysis

Streamlit
Plotly
Aggregates-only AI
HTML report

Exploration plus written explanation—not KPI tracking: one business CSV yields profile, a fixed chart pack, rule-based quality hints, and optional OpenAI narrative built only from aggregates (never raw rows), bundled as one HTML story for readers who need context, not a metric wall.

GitHubLive demo

Interactive demo

Forecasting app

Streamlit
Holt-Winters
Plotly
Rolling backtest

Univariate demand curves with honest baselines—not neural nets: resampled Holt-Winters by day/week/month only when seasonality is supportable; Plotly band with MAPE/MAE against naive; warnings and fallback reasons stay visible for operators validating the forecast.

GitHubLive demo

ML systems · implementation references

APIs, batch jobs, retrieval, and feature pipelines (featured ML entries are above).

ML implementation

RAG document intelligence QA

FastAPI
FAISS
sentence-transformers
Docker

Answers questions from your own documents in plain language and points to the exact source for each answer.

Ingest chunks documents for retrieval, then answers questions using retrieved context and a hosted model.

GitHubLive demo

ML implementation

Train–score feature compilation pipeline

pandas
pytest
FastAPI
Batch pipeline

Keeps training and downstream scoring aligned on the same engineered fields so exported rows match what the model was fitted on.

A batch compilation step turns approved raw extracts into a versioned feature table under locked column contracts.

GitHubLive demo

Experimental & internal buildsOptional—viewers and tooling outside core storefront offers. Collapsed by default.

Experimental reference

ML monitoring & batch data quality

pandas
pytest
Streamlit
PSI / KS

Shows when freshly scored batches or inputs drift away from an agreed baseline so teams react before KPIs silently rot.

Compares each new batch against a frozen reference distribution and emits structured summaries for inputs, categorical shifts, and score behaviour.

GitHubLive demo