Data pipelines & ETL
I connect your scattered sources — APIs, spreadsheets, databases — into one clean, scheduled flow that lands where you need it, on time, every time.
$ automation.status() → active
Python automation and machine learning for teams tired of doing the same task twice. I design, build, and ship the systems that quietly do the work — so you don't have to.
Replace these with your real numbers before publishing.
# what I automate
I connect your scattered sources — APIs, spreadsheets, databases — into one clean, scheduled flow that lands where you need it, on time, every time.
From forecasting to classification, I build models that hold up outside the notebook — trained, tested, and wired into something your team actually uses.
The repetitive clicks, copy-pastes, and status-checking your team does by hand — I script it away and hand you back the hours.
Two systems that should talk to each other but don't — I build the bridge, so data moves without anyone re-typing it.
# how it gets built
We map the current process end to end — what's manual, what breaks, what it's costing you in hours.
I sketch the pipeline: inputs, transformations, decision points, and where a model earns its place.
Python, tested as I go. You see working pieces early, not a black box at the end.
Scheduled, containerized, and monitored — running on its own, not on someone remembering to click "run."
Alerts on failure, logs you can read, and a handover doc so the system outlives the project.
# selected work
Logistics · ETL
Replaced a nightly manual spreadsheet merge with a scheduled Python pipeline reconciling stock across systems.
Retail · ML
Built a forecasting model that cut overstock without increasing stockouts, retrained automatically each week.
SaaS · Workflow
A classifier routes and tags incoming tickets before a human ever opens them, with a fallback queue for edge cases.
# who's building this
I'm a Python developer who specializes in automation and machine learning — the layer of work between "we have the data" and "the system handles it." I've spent my career turning manual, error-prone processes into pipelines that run quietly in the background, and models that hold up once they leave the notebook.
I care less about the fanciest algorithm and more about whether the thing still works at 3am when nobody's watching it.
# let's talk
Tell me what it is. I'll tell you honestly whether automating it is worth it.