Aniket Apte is building in the AI era.
A data professional with an MBA from Chicago Booth, fluent in Python and the language of decisions. The leap from analyzing the work to shipping it is shorter than it has ever been. This is an honest map of what is already underway, and the next move.
The launch is already underway.
01 / In the openAn ai_agent built from the ground up
Wiring an LLM into a loop of tools and actions. The exact pattern that turns a model from a chat box into something that does the work.
A static_site generator, by hand
Turning content into a deployed site from first principles. Knowing how the machinery works is what makes the next build fast.
A Django mvp-landing + bookbot
A working web app and a text-analysis CLI. The full arc from a single script to something that runs end to end.
Studying the new agent tooling closely
Notes and references on how modern AI coding agents are actually built. Learning the frontier from the inside, not the headlines.
Honest note. Every name above is a real, public repository under github.com/aaptedata. Several began as guided coursework, that is exactly the point. The hard part of any field is not the first lesson, it is the next ten miles. The momentum is real and the foundation is in place.
Where a data mind goes next.
02 / The on-rampTurn analysis into agents
You already know how to find the answer in the data. The next step is wrapping that judgment in a small agent that runs it on its own, on a schedule, without you in the loop.
from notebook to running toolShip faster than you can spec
With an AI pair, the distance from idea to a deployed, working thing collapses. The MBA framing and the Python both compound: you decide what matters, the agent handles the rote build.
spec in the morning, live by nightCompound in public
Each build is a commit, each commit is a record. Working in the open turns scattered effort into a visible body of work, the thing recruiters, collaborators, and your future self can actually see.
proof, not promisesThe best time was years ago. The second best is the next commit.